Safe In-Wall AC to DC Transformers??

[rvendrame]

Adding my two cents, I have dissected two devices (one zwave from GE and one unbranded, non zwave) and both used a capacitor-resistor power supply.

Very dangerous, as it is "transformerless" --- In other hand, cheap and small footprint. By using X2 capacitors the risk is a bit lower. It is in my list for prototype/testing over next weeks.

[sundberg84]

@rvendrame So did i on a 433mhz relay sold quite common here... it looked as this capacitor-resistor power supply as well.

[petewill]

@rvendrame said:

Adding my two cents, I have dissected two devices (one zwave from GE and one unbranded, non zwave) and both used a capacitor-resistor power supply.

Very dangerous, as it is "transformerless" --- In other hand, cheap and small footprint. By using X2 capacitors the risk is a bit lower. It is in my list for prototype/testing over next weeks.

Interesting. Thanks for posting. I have very limited knowledge in this area so I'm learning a lot. I'm excited to get a good solution for this. It will be very useful for many things.

[Moshe Livne]

@rvendrame very interesting. The link mention only 100ma at most but I guess that they are using SSRs so don't face the problems we have with our relays. I wonder why there is no "off the shelf" unit like that. I would be a bit scared to construct something like that.

[Moshe Livne]

@ServiceXp @petewill @Didi here is another one! http://www.farnell.com/datasheets/606238.pdf

this one is reasonably smallish, has all the required protections. They are so well hidden....

[Didi]

@Moshe-Livne said:

@ServiceXp @petewill @Didi here is another one!

And pice is ok

Found one at ebay:
http://www.ebay.de/itm/5V-550mA-switching-power-supply-module-MYRRA-47122-PCB-mount-/301663807573?pt=LH_DefaultDomain_3&hash=item463c906855

[rvendrame]

@Moshe-Livne , I believe there is no off the shell as if you dig in the details, you will see that the impedance of the consuming circuit ( arduino + radio + relay in our case) direct influence the overall performance of the capacitor/resistor model.

Usually triacs consume less than relay (something between 10 and 50mA), and you can also control the bright of a light or speed of a motor, instead just turn them on or off.

The downside, you need a snubber circuit as well as a zero-cross detection circuit. The snubber is also recomended for relays, though.

So at the end you don't reduce the footprint or costs --- But triacs tend to last longer than relays, as no mechanical commutation occurs.

[DrJeff]

@rvendrame I'm right with you on all of this have been trying to do this for a couple months with no time right now. Actually it has been longer I was originally building all this for another system using XBees and then found it better to use esp8266 then said better yet use Mysensors so here we are! Transformer less is key at least I think so! Only problem no off the self builds.

[scalz]

Hi,

I found an application notes from enocean regarding power supplies design. I think it could be helpful, and good to remember, so I share the link.
http://www.enocean.com/fileadmin/redaktion/pdf/app_notes/AN101_POWER_SUPPLY_LAYOUT_2015.pdf

[rvendrame]

I'm currently studying this doc STEVAL-ILL004V1 AN2263 AppNote .pdf. I should prototype something over next week if family allows.

[scalz]

Hi. interesting link, I will read it when I will have time. It looks like u2008 or u2010 which are obsolete now (I have some in stock, cheap on aliexpress) but I didn't know your ref. thanks

[rvendrame]

@scalz, I'm trying to porting the capacitor-resistor power supply concept from that doc to our mySensor world. Specially for 1-wire dimmer, as described in chapter 3.1.

[scalz]

@rvendrame : I tried something similar once. I didn't remember exactly how all was connected but with triac, in one wire, even if I was not toggling anything, I had brown reset at start (maybe I could find my schem). With two wire, I had no problem but a little flickering (zero cross detection done by arduino). Maybe it was that which restarted. Next, I did another test with mosfet pwm, in one wire config again. This time, it started ok, and as I was increasing pwm it restarted for pwm = 170-200 (I don't remember exactly) .But I jumped onto another project I am finishing.
I will follow your results with interest as your link is interesting , and at my home it is one wire too.

[rvendrame]

@scalz, with the 1-wire circuit , the maximum you can reach is like 80% - 90 % of light bright, as you need to keep the triac/mosfet "off" for a short period, in order to recharge the PSU capacitor. Maybe that explained why it re-started on your test at PWM 170-200. (and would also confirm that it works 'as designed', good to know)

[ThinkPad]

I would be careful when stripping cheap USB-powersupplies and tucking them into a box in the wall. Read this page: http://www.lygte-info.dk/info/indexUSB UK.html and this page: http://www.lygte-info.dk/info/usbPowerSupplyTest UK.html to see how badly constructed some of these powersupplies are.

He also explains how he is testing those powersupplies: http://lygte-info.dk/info/usbPowerSupplyTestHow UK.html
Very informative i must say.

[Moshe Livne]

@ThinkPad Fascinating read.... If you read through the thread you will see that we actually go to the same conclusion. The meanwell charger should be good - they are respectable company and it has the 3 safeties we were looking for. It is not a very common charger - I have never seen it being sold before.

[petewill]

@ThinkPad Thanks for posting. It's good to see how the tests are performed. Unfortunately I don't have the equipment or skill to know how to test. Hopefully we will be able to find a safe one that we can all use.

@Moshe-Livne have you been able to find the meanwell chargers from anywhere besides jameco?

[Yveaux]

@petewill Mouser carries them, and also the european version (GS05E-USB). Please see here.

[Sparkman]

@ThinkPad Thanks for posting that. Great info there. Looks like you get what you pay for with the cheap power supplies.

Cheers
Al

[Didi]

@ThinkPad said:

I would be careful when stripping cheap USB-powersupplies and tucking them into a box in the wall.

Thanks for the information.The most powersupplies are horrible.
The engineers have developed them must go to jail.
Safety must be the major goal not the price.