Y capacitor question - how to make this SMPS safer?

[sparky.summer]

Greeting to all,

Recently I bought an AC-DC 24V/4A SMPS module on ebay (pictured below), which I plan to use with my T12 Soldering Iron. There is a Y1 capacitor connecting the output negative to the mains (through the diode bridge), as already described in another post. The SMPS module is a 2 wire type, i.e. there is no Earth connection on the PCB, so I guess this makes it Class 2, Unearthed PSU.




Being the kind of person who wouldn't stake his life on a Chinese Y cap, I decided to do a little redesign, but I am not a PSU or EMI expert, so any comments and suggestion will be highly appreciated (and will probably prolong my days on Earth, if correct).

The plan is to disconnect the Y capapacitor from the Live part of the circuit and wire it to Earth on the power plug. Two things to note:

1. The Soldering Iron tip is already connected to Earth through 1.5M wirewound resistor.
2. The Earth connection on the AC outlet is actually connected to the Neutral wire.



Do you think that this reconnection of the Y capacitor will increase the safety of the SMPS or not?

[Ian.M]

Assuming your wall outlet is coimpliant with your national electrical code (regulations), then the Eatth pin is intended for the safe grounding of exposed metalwork on appliances.   The pri-sec class Y cap is *required* in most types of two wire SMPSU to shunt HF switching leakage current capacitively coupled from the primary to the secondary back where it came from.  It should not be removed, however replacing it with one you know isn't fake would be sensible if you are keeping the PSU ungrounded.  However it would make far more sense to simply ground the secondary side 0V rail.

Changing it to a secondary to earth cap doesn't return the HF leakage current back where it came from unless you add a delta filter network  between L, N and E at the input of the PSU.

[ovnr]

And if you do ground the secondary side directly, please remember it if you try to solder on a ground-referenced circuit that's live. Because you'll be shorting whatever you touch with the soldering iron to ground.

[Zero999]

What makes you think the Y capacitor is fake and doesn't comply with the relevant safety regulations?

If the capacitor is really dodgy, then the transformer won't be any better, so you might as well replace the power supply with one which you're sure is safe.

Rather than removing the Y capacitor. If you don't want to connect the output directly to earth, you could add another capacitor between the negative rail and earth. This will divert the tiny AC leakage current which can give you a tingle back to earth, yet keep it separated from earth at DC, without compromising the EMC.

[sparky.summer]

Thank you all for the answers!

I am never going to solder live circuits. I want to make the Soldering Iron ESD safe and also human safe. At the same time I prefer not to connect the tip directly to Earth, if possible.

Do you think that this second redesign is a better solution?

[Ian.M]

What handpiece are you actually using?
Does it have separate connections for tip grounding and the heater element?
If the tip ground is common with one side of the temperature sensor or heater element, then Hero999's suggestion is probably best.  If the tip grounding is seperate from the sensor and element, then you can ground the PSU 0V without having to hard-ground the tip.

[sparky.summer]

Thanks for the Suggesion, Ian.M
The tip is separate - not connected to the heating element/thermocouple. There is another 1M resistor on the soldering iron controller, connecting 0V to tip's EARTH. Should I Earth the tip for ESD safety or not and if yes, how?

[Zero999]

Thank you all for the answers!

I am never going to solder live circuits. I want to make the Soldering Iron ESD safe and also human safe. At the same time I prefer not to connect the tip directly to Earth, if possible.

Do you think that this second redesign is a better solution?

No, don't connect the 0V to neutral. That's very bad because the neutral can become disconnected, then the circuit will be live.

Use a capacitor to connect 0V to earth at AC.

[aandrew]

Interesting discussion.  Now a question -- why 10nF? why not another value? They used 2.2nF between secondary 0V and neutral, presumably because of the Xc at the switchmode regulator's switching frequency.

[Ian.M]

Thanks for the Suggesion, Ian.M
The tip is separate - not connected to the heating element/thermocouple. There is another 1M resistor on the soldering iron controller, connecting 0V to tip's EARTH. Should I Earth the tip for ESD safety or not and if yes, how?

If you earth the 0V rail, the tip will be ESD safe as it will be earthed via that 1Meg resistor.

However, the heating element ceramic insulation may leak current from the element to the tip when hot, especially if the element becomes grossly contaminated with burnt flux, so you should check the DC voltage across the 1Meg resistor, with the iron hot, turn up the temperature to make sure the element's on, and record the voltage, then check it again at regular intervals.   If it starts to increase excessively and you don't want to replace the element, you may be able to carry on for a while by reducing the resistor, but eventually, you may have to hard ground the tip.  A volt or two is tolerable as with only a couple of uA of leakage driving it there is minimal chance of damage even with sensitive devices, but if you work with unbuffered CMOS or unprotected small MOSFETs, any larger voltage could be problematic.

[Zero999]

Interesting discussion.  Now a question -- why 10nF? why not another value? They used 2.2nF between secondary 0V and neutral, presumably because of the Xc at the switchmode regulator's switching frequency.

Nothing to do with the switching frequency.

The secondary side will normally float at half the mains voltage, which will be 120V on a 240V supply.  10nF is 4.54 times the value of 2.2nF, forming a 1:5.54 capacitive potential divider, reducing the voltage to 21.6V, which you won't be able to feel.

I don't know the value of the Y capacitor in this case but 10nF is probably a decent enough value to connect the secondary to earth.

[Gyro]

Thank you all for the answers!

I am never going to solder live circuits. I want to make the Soldering Iron ESD safe and also human safe. At the same time I prefer not to connect the tip directly to Earth, if possible.

Do you think that this second redesign is a better solution?

Actually, strange as it may seem, you do want to connect the tip directly to earth if you want it to be human safe. If you accidentally try to solder something that is still at high voltage (charged capacitor or something you've forgotten to disconnect) an earthed soldering iron tip will ground it and protect you. If you ground the tip via a resistor, then the first thing you might know about it is when you touch the joint with the solder. In extreme cases, the tip would become sacrificial, but it could save you from electocution!

[Ian.M]

A hard-grounded tip can get you into serious trouble if you are working on boards with soldered in batteries and any external connections that bring in a ground.  However it is the usual arrangement, and as long as you disconnect (not just power off) any board you are working on, doesn't cause any issues.

[alm]

There are indeed some (for most people) rare cases where a grounded tip complicates things, but in the majority of cases it improves safety, and it seems to be the norm in most soldering stations I have seen.

[Ian.M]

A hard-grounded tip is an essential safety requirement for any soldering iron with a mains voltage heating element.  However, if the iron has a low voltage element, the only reasons to ground the tip are to minimise the risk of ESD and to minimise the leakage current from the element to the device being worked on.  Grounding the tip of a low voltage iron does NOT improve safety, unless it is being operated from an unsafe PSU.   In fact it increases the hazard by introducing a ground into what may well be a floating piece of equipment, and ensuring that when it matters, the user is holding a hot object in one hand and a highly conductive length of solder in the other with its end already in close proximity to whatever part of the board has stored charge.