Switching mains AC power with small switch

[VEGETA]

Hello,

I have a design of power supply which takes 100-220v AC. It has a button which must be in a specific location (it is a replacement of Sony Playstation 1 internal PSU) and I am not comfortable using it in-line in the power circuit. On the other hand, I still require the circuit to be fully off until the switch is pressed.

therefore, I thought of this circuit:

AC mains -> the small switch -> bridge rectifier -> resistor divider (using regular resistors) -> small linear regulator -> solid-state relay ---> which outputs the mains to go into the desired circuit.

this way, I have the device fully off until you press the button. Then the divider will get the voltage to about 10-40v or so which feeds into the linear regulator. this one delivers something like 3.3v or less to activate the solid state relay.

here the are the parts I settled on:

solid-state relay: Panasonic AQH2223 ---> requires 1.21v input voltage, I assumed that I cannot deliver 3.3v therefore a small limiting resistor of 100 ohms can work right? or use a 1.25v linear regulator?
linear regulator: AP7384 or TPS7B6933
bridge rectifier: ABS10A-13

notice that this activation circuit will be very small current, basically < 10 mA or so. thus, I assumed any cheap switch can do the job since 220v AC with such low requirements can be ok.

what do you think of this?

any other suggestion?

[calzap]

Is the small switch listed (UL, ETL,TUV or etc.) and rated for 220 VAC?  An input fuse should be used.

Mike

[VEGETA]

Is the small switch listed (UL, ETL,TUV or etc.) and rated for 220 VAC?  An input fuse should be used.

Mike

i still didn't choose the switch but it will be rated for high voltage... still cheap from LCSC or my supplier in Alibaba.

so you suggest something like 20 mA fuse for this activation circuit?

assume the switch is ok, is the circuit mentioned good?

what about the panasonic solid-state relay input? you suggest my choice of 3.3v + resistor in series or just use 1.25v regulator?

regards

[calzap]

If the circuit draws 20 mA, you’ll want a higher rated fuse.  Probably something like 100 mA.  Might wait until everything is laid out and measure actual current draw before deciding on a fuse.  Be careful testing low current fuses with a multimeter.  The meter may produce enough current to blow them unless you use a series resistor.

I won’t comment on other components without seeing a schematic and might not then.

Mike

[xvr]

You can use something like that https://www.boospa.net/en/tbs106a-pneumatic-switch.html for your main power switch.

[VEGETA]

If the circuit draws 20 mA, you’ll want a higher rated fuse.  Probably something like 100 mA.  Might wait until everything is laid out and measure actual current draw before deciding on a fuse.  Be careful testing low current fuses with a multimeter.  The meter may produce enough current to blow them unless you use a series resistor.

I won’t comment on other components without seeing a schematic and might not then.

Mike

please check suggested schematic.

I could add the fuse later, probably best suited after the rectifier.

the main ac input feeds the bridge rectifier, then with resistor voltage divider it delivers 9-22v dc to LDO. LDO easily delivers 3.3v and with limiting resistor it ensures small current to the SSR which has 1.2v drop at the internal LED. Then it delivers the AC_L power line to the target circuit while AC_R line is connected to it directly.

this is what I have suggested.

these panasonic relays are interchangeable so that if one is not available at lcsc i can get another one from digikey or so. the smallest one of 300 mA is enough since PS1 is rated at 10W (or 17W in some extreme cases). this makes 100v AC at 17W requires 170 mA, most cases will be a lot less than this.

I wished for slightly cheaper solution but can't figure it out now.

You can use something like that https://www.boospa.net/en/tbs106a-pneumatic-switch.html for your main power switch.

please check the image attached.

the image is the power supply i want to create a substitute for. it has to have the showed button in that exact place and exact form in order to fit and work properly.

I could not find such buttons to be rated for higher voltage till now, except in LCSC which I don't know if I can trust it xD... but this one for example can work in the activation circuit since it won't need much current: https://www.lcsc.com/product-detail/Push-Switches_XKB-Connection-XKB8080-Z_C318860.html

the original thought was to get one which can directly switch the ac power, that one requires say 200mA in the same formfactor as seen in picture below. i could not find one.

and my requirement is to have the entire circuit totally powered off when the button is not pressed... so i can't for example have the bridge rectifier + LDO before the switch in order to use low voltage switch. this way they will be active all the time.

[xvr]

Take a look at my suggestion. It completely mechanical and did not required any power to operate. It rated to 220V 16A. It is pneumatic operated switch. It's button can be install in any place, switch itself unstalled separately. They connected by pvc tube. You just have to find pneumo button of desired shape.

[VEGETA]

Take a look at my suggestion. It completely mechanical and did not required any power to operate. It rated to 220V 16A. It is pneumatic operated switch. It's button can be install in any place, switch itself unstalled separately. They connected by pvc tube. You just have to find pneumo button of desired shape.

the one you linked is very robust for sure, but its price is more than my entire BOM xD. I couldn't find something like it on digikey.

I found stuff like this: https://www.digikey.com/en/products/detail/e-switch/PB1973ABLKBLKEF0/1426844

but this one is also expensive and its height may not be compatible.

right now i am asking my alibaba button\switch manufacturer if they have a catalogue of compatible parts. I am getting some stuff from alibaba anyway, so adding another part is not a problem.

I am still interested in opinions about my circuit above... since it does the job using regular cheap lcsc parts with no danger. i just prefer a simple button though.

[xvr]

I couldn't find something like it on digikey.

These kind of switches used in bath and kitchen equipment. You can try to search something in spare part for them (jacuzzi, food waste disposer, etc)

I am still interested in opinions about my circuit above...

It could be made simpler with capacitor dropper. Basic circuit in attachement.
It will replace all your circuit from button and till R10 & U5

[VEGETA]

I couldn't find something like it on digikey.

These kind of switches used in bath and kitchen equipment. You can try to search something in spare part for them (jacuzzi, food waste disposer, etc)

I am still interested in opinions about my circuit above...

It could be made simpler with capacitor dropper. Basic circuit in attachement.
It will replace all your circuit from button and till R10 & U5

will this circuit you suggested be totally off (no current drawn at all) when button is not pressed?

[xvr]

will this circuit you suggested be totally off (no current drawn at all) when button is not pressed?

Shure. It will be power via button, plugged in series in mains power. There is no current path when button is not pressed.

[VEGETA]

will this circuit you suggested be totally off (no current drawn at all) when button is not pressed?

Shure. It will be power via button, plugged in series in mains power. There is no current path when button is not pressed.

so the button is placed before R1 in your schematic?

what would the final dc voltage be? since I need it very small for the panasonic SSR. at least in a range acceptable for the LDO. since I still need a SSR to activate the power.

[xvr]

Yes, button placed between AC mains and R1. Output voltage will be any you like - it defined by zenner voltage. And it will be regulated - you don't need LDO.

[xvr]

R1 limit inrush current - it can be a small one. Main voltage drop performed by C1. It shoud be rated for AC voltage. C2 perform output voltage smoothing.

[VEGETA]

R1 limit inrush current - it can be a small one. Main voltage drop performed by C1. It shoud be rated for AC voltage. C2 perform output voltage smoothing.

so how much capacitor value?

quick search i found this: https://www.lcsc.com/product-detail/Multilayer-Ceramic-Capacitors-MLCC-SMD-SMT_PSA-Prosperity-Dielectrics-FM43X104K631EFG_C394157.html

but it is just 100nF.

ceramics won't be of huge values at mains voltage, so am i required to get elec. caps or polymers? these are polarized tho

[xvr]

Let's set 10mA current in capacitor net. For 220V AC amplitude will be 220*✓2 = 308 Let take 10% of possible voltage drop. It will be around 280V. So we need impendance of 280V/10mA = 28K
Capacitance impendance is 1/2*pi*f*c ~ 1/314*c
Here C is 0.1uF (just your capacitor)
Capacitor should be film capacitor (not polarised) MKP or MKT or analog

[VEGETA]

Let's set 10mA current in capacitor net. For 220V AC amplitude will be 220*✓2 = 308 Let take 10% of possible voltage drop. It will be around 280V. So we need impendance of 280V/10mA = 28K
Capacitance impendance is 1/2*pi*f*c ~ 1/314*c
Here C is 0.1uF (just your capacitor)
Capacitor should be film capacitor (not polarised) MKP or MKT or analog

so can i for example put 3x in parallel to have 0.3uF total and still use ceramic? i guess x and y caps can't work here... or if they do, then they are readily available.

as for film caps, here are my findings:

https://www.lcsc.com/product-detail/Film-Capacitors_JOEY-MEF104K2G3AS3040A1_C18295884.html  --> can put 2 or 3 in parallel since it is very cheap

https://www.lcsc.com/product-detail/Film-Capacitors_Vishay-Intertech-BFC237357104_C3778615.html --> from Vishay but a lot more expensive.


those are 400v, going down to 250v can be cheaper.

[DavidAlfa]

So much complexity... just use a small relay?

[xvr]

Yes, you can use ceramic (but check datasheet for conditions - ceramic of such voltage could be very special). You can also use X and Y capacitors, but they could be bulky.

> as for film caps, here are my findings:

Yes, they will fit

> So much complexity... just use a small relay?

Relay with 220AC coil will be not small 😕

[VEGETA]

So much complexity... just use a small relay?

this is good idea but i am trying to stay away from any mechanical part... will consider this if all solutions failed.

still, i would need regulation and step down for voltage in order to be in the 5v range of such small relays... those with full ac coil rating are a lot more expensive, 5$ or so in lcsc. the SSR i have is kinda ok in terms of price.

Yes, you can use ceramic (but check datasheet for conditions - ceramic of such voltage could be very special). You can also use X and Y capacitors, but they could be bulky.

what type of ceramics are used here?

x and y caps can be bulky but i do have some space... the space i have is about 20mm width and 30mm length.

here are some options for x and y caps:

https://www.lcsc.com/product-detail/Suppression-Capacitors_KEMET-R463I310050M1K_C3298492.html  --> 100nf and cheap if only one is needed, even if using 2 in parallel is also cheap.

https://www.lcsc.com/product-detail/Suppression-Capacitors_TDK-B32921C3473M189_C125424.html --> 3 in parallel for 141 nF, still cheaper than 2x of the above one. 3 of it can fit in the space mentioned. this one is 305v though.

those are Kemet and TDK brands, if I go with Chinese brands it will be a lot cheaper. if it is gonna be reliable then 3 of

this one

can cost a fraction of all above.

do you recommend going wit 3x in parallel or just one?
can chinese parts (especially in parallel) be reliable here?

[DavidAlfa]

Or leave the primary rectifier connected (Won't consume any power) and use a fet to switch the output on/off.
BTW always use a PTC to limit capacitor inrush current, otherwise charging currents will get huge and stress any device switching it (Fets, relays, switches) causing early failure.

I know this circuit looks "ugly" because it's left floating, but its mains anyways, everything there is dangerous and should be completely isolated from the secondary.
Omitted the obvious fuse !

[xvr]

Any ceramic capacitors here should be checked first with they datasheet. There is no any generic recommendations here. Ceramic of such voltage are always unique.
As for X capacitors - 1 pcs preferably than multiple in parallel, at least from reliability point of view. TDK and Vishy also preferable. As for Chinese - take a look in their datasheet. If they do not have one it will be a clear signal do not use them. Check for datasheet completeness first, for characteristics second.

[VEGETA]

Or leave the primary rectifier connected (Won't consume any power) and use a fet to switch the output on/off.
BTW always use a PTC to limit capacitor inrush current, otherwise charging currents will get huge and stress any device switching it (Fets, relays, switches) causing early failure.

I know this circuit looks "ugly" because it's left floating, but its mains anyways, everything there is dangerous and should be completely isolated from the secondary.
Omitted the obvious fuse !

so you used a 15v zener with those 1M + 100K resistors to put 15v at N-Mosfet gate... then it will be activated by the switch. when On, the circuit gets complete and secondary side is fully on.

then i can use that high voltage dc and step it down + regulate it to be able to supply the necessary voltage to SSR right?

Any ceramic capacitors here should be checked first with they datasheet. There is no any generic recommendations here. Ceramic of such voltage are always unique.

i found tons of them in LCSC which are X7R or X7T. I guess since they are in ac voltage series circuit, they won't suffer from DC bias.

here are some good ones I found:

https://www.lcsc.com/product-detail/Multilayer-Ceramic-Capacitors-MLCC-SMD-SMT_TDK-CGA5L4X7T2W104MT0Y0N_C2991839.html
https://www.lcsc.com/product-detail/Multilayer-Ceramic-Capacitors-MLCC-SMD-SMT_PSA-Prosperity-Dielectrics-FM31X104K451EEG_C784091.html

maybe i forgot to inform you that the power supply will be in 100-240v 50-60hz range. you calculated 113 nF to be ok for 220v AC 50hz, it gave 28k so the same thing for 100v ac would be 14k impedance. that is without considering the slightly higher amp consumption if needed.

therefore, wouldn't it be a better choice to use 3x parallel of those ceramic caps listed above? they will give 300nF and will be robust since they are regular ceramics, not x or y rated.

what do you think?

BTW, voltage rating should be taken as rms or peak? like there are 275v rated x caps... in our case of putting them in series, what should we consider the voltage be?

[xvr]

then i can use that high voltage dc and step it down + regulate it to be able to supply the necessary voltage to SSR right?

I think DavidAlfa means to power device from this voltage directly, without any SSR

what do you think?

TDK part looks good. Prosperity Dielectrics also looks not bad.

maybe i forgot to inform you that the power supply will be in 100-240v 50-60hz range. you calculated 113 nF to be ok for 220v AC 50hz

It is ok with a quite marging. 10mA of output should be enough for both zener and SSR. 5mA also could be enough. Just check how much current SSR required. And what minimum regulation current of zener.

[VEGETA]

I think DavidAlfa means to power device from this voltage directly, without any SSR

the output of his circuit is high voltage dc though. I still need something to connect the mains ac to my target.

TDK part looks good. Prosperity Dielectrics also looks not bad.

nice, then I could use 10x in parallel to have 1000nF total (or other smaller solution like 5x220nF or just 1uF cap alone), should be good enough for all cases since 1uf impedance is 3.2k and 2.65k for 100v ac and 220v ac respectively which ensures more than enough current is there since for 100v it will be around 31mA/37mA and for 240v will be 75mA/90 mA for 50/60 hz respectively. won't cost much too. good that cheap reliable ceramics can work.

here is official datasheet for SSR: https://api.pim.na.industrial.panasonic.com/file_stream/main/fileversion/2787

If I understand correctly, it needs 20 mA forward current for the LED at 1.21-1.3v but under that it is mentioned that trigger LED current is 10 mA and yet under that the recommended LED current is 15-20 mA, so we assume 20 mA... therefore I could use 5 or 3.3v rated zener (or LDO) then put a series resistor to limit the current.


this Zener seems ok: https://www.lcsc.com/product-detail/Zener-Diodes_Diodes-Incorporated-BZT52C3V3-7-F_C151564.html

it is 3.3v, with a series 100 ohm (or 120 ohm) resistor, combined with 1.21v (to 1.3v max) SSR internal LED drop it can deliver 20 mA or near that... certainly within the range of 15-25 mA recommended. chosen resistor must be 250mW to ensure no issues.