0
\$\begingroup\$

I have been perplexed recently with regards to people on a forum, stating that it is unsafe to put 240vac directly into a basic potential divider. The divider was literally 210k ohms, with 3 resistors. They were 100k, 100k, and a 10k resistors. The usable part to which voltage was measured from was the 10k, and utilized through a basic PSU set-up (diode bridge and smoothing cap). The math is obvious, and no need to mention that here.

I have done research in this type of thing, where one is essentially bypassing the use of a transformer. I have looked at the pros and cons etc, and they are easily understood. I have even come across small portable appliances that use this set-up as opposed to a transformer e.g. a very light mains powered wireless doorbell. The transformer would add an undesirable weight, because it just plugs directly into a 3-pin UK wall socket.

My potential divider was only a prototype, and simply for testing various aspects, and to re-familiarize myself with the concept of a power supply set-up (due to a gap in the last time I worked in a professional position in industry). It was not meant as some final product at all. Far from it.

enter image description here

I cannot see any health and safety concern other than the obvious ones associated with dealing with mains supply (e.g. snagging cables, or putting something like a measuring device on the side of the table, falling off, and have it pull the whole set-up into your lap!)

I am fairly certain I am not missing something here. If anyone has some advice regarding this type of set-up, and whether it should have some device preceding the 240vac in, then please enlighten me. I would like to know.

Just as some additional information, I am aware that putting a physical fuse into the circuit would avoid simply using the mains panel CB to trip. In other words, a fuse set before 240vac in would indeed add additional protection, so one is not ONLY relying on the CB.

\$\endgroup\$
14
  • \$\begingroup\$ @user1937198 Is not the likelihood of the 10k resistor failing somewhat negligible? Also, if it failed, would that not more likely just mean that it could go open circuit (I believe, based on experience they are more likely to be OPEN). Even if it shorted, then there is still 200k ohms in the circuit, so there is still no apparent problem. Obviously one takes these things into consideration. \$\endgroup\$ Commented Jul 24 at 12:57
  • \$\begingroup\$ It may look odd because it appears that the user removed the comment. That is why the previous comment is there. \$\endgroup\$ Commented Jul 24 at 13:01
  • 2
    \$\begingroup\$ The output is not ELV as it is referenced to the mains. The other thing to consider is transients due to switching and lightning. These can easily exceed the stated mains voltage which is why larger creepage and clearance is needed. A fuse is not going to stop you being electrocuted. You can’t rely on neutral being 0V as it is fairly common for the neutral to be lost and you get the phase to phase voltage. \$\endgroup\$ Commented Jul 24 at 13:48
  • \$\begingroup\$ @Kartman Yes, of course the fuse does not stop electrocution. The fuse would be stop stop a short, and avoid relying on the CB. I have noticed that CBs operate well before the fuses in the 3 pin UK plug. WAY before they blow. The fuse could be super low current rating, considering that the current is literally milliamps. Can you elaborate on why you say it is not ELV? To the best of my knowledge, the voltage out at that point is ELV due to it being well below 50v ac. \$\endgroup\$ Commented Jul 24 at 13:56
  • 3
    \$\begingroup\$ It can’t be legally classed as ELV as it’s not isolated from the mains. For the reasons I previously noted, your ‘ELV’ could easily not be as low voltage as you hope for. \$\endgroup\$ Commented Jul 24 at 14:02

2 Answers 2

8
\$\begingroup\$
  • This circuit does work.

  • If the 10k resistor fails open (and resistors do normally fail open), nearly-full 240V goes to the output (or a higher voltage depending on load), probably blowing it up.

  • Both output wires should be treated as live wires, for the purposes of safety. You may think that by connecting the output near neutral, they don't have to be treated as live, but consider that if someone brings your circuit from the UK to continental Europe and plugs it in with an adapter, live and neutral may be reversed.

  • Because of the above, this can only be used if everything downstream of the power supply is insulated. It's not clear what "a doorbell" means to you. If this is going to connect to a doorbell button with wires (you mention replacing a doorbell transformer) that's not safe because the whole doorbell system would have to be treated as live wires. If it's all contained in a plastic box, and you're just receiving a wireless signal and playing a dingdong on a speaker, it could be okay.

  • Also you must not let them touch anything else earthed. Measuring the output on one channel of your oscilloscope carries a risk of blowing a fuse (or worse) on your oscilloscope. A differential measurement is required, since the output isn't reliably earth-referenced or floating. And of course, avoid touching the things you're measuring.

  • Each 100k resistor dissipates about 130mW continuously. Make sure to size them adequately. A typical 250mW resistor may be technically able to handle 130mW, but it will get rather hot, possibly hot enough to melt some plastic. This is why, when a voltage divider circuit is used like this, a capacitive divider is preferable to a resistive divider.

    Burning under a watt, continuously, isn't that much power, but it can add up if you sell a million of these things and they're plugged in all the time, especially because that power is serving absolutely no purpose. There've been initiatives in different countries to try and reduce standby power.

  • Capacitive voltage dividers are used in real products (cheap ones) so evidently, it's acceptable to use this type of design in the real world. Such products are double-insulated so nobody can poke any of the circuitry inside and get electrocuted, even if live and neutral are reversed. The principle of a capacitive divider is pretty similar to what you have here, except with capacitors, and it only works on AC.

    Just like a resistive divider, the capacitor size should be chosen to provide the needed output power without too much idle current. Unlike a resistive divider, the current is out of phase with the voltage so the unused energy is returned to the mains each cycle instead of being turned into heat.

  • Compliance requirements are out of scope of this answer because I don't know anything about compliance.

\$\endgroup\$
3
  • \$\begingroup\$ OK, got it. The main point was that it was simply a very basic prototype thing. Apparently the immediate concerns are minimal, especially with the experience I have regarding the math and science of the set-up. If one were to actually produce a workable device, one would follow "compliance", like what you stated in the last bullet point. \$\endgroup\$ Commented Jul 24 at 13:22
  • \$\begingroup\$ And also, with the set-up, the 10k failing does not seem to present a "major" problem due to the fact that there is still the other 200k resistance in the circuit. That was also why I put that safeguard with using 2x 100k, as opposed to simply a 200k. If the 10k shorted, there simply would not be anything at all (unless I have overlooked something!) \$\endgroup\$ Commented Jul 24 at 13:27
  • \$\begingroup\$ Oh stupid me! Of course there would be a voltage sorry. Oh dear, I am really out of practice. I am just going to delete the comment that preceded this one to avoid embarrassment. \$\endgroup\$ Commented Jul 24 at 13:39
3
\$\begingroup\$

You have used the term ‘ELV’ in the diagram. Not a good idea, as that has a specific meaning in some jurisdictions (and it is likely wrong).

In fact both sides of that output require the same very reliable high-voltage insulation from any human contact as the 240VAC mains itself, due to the risk of current flowing through a user to earth ground. Typically this will be “reinforced” insulation (sometimes called double insulation). Human danger and severe damage from a similar cause could also result of this circuit was connected directly to some other circuit such as an oscilloscope or computer.

You could use an isolation transformer to provide galvanic isolation, however they are generally heavy and expensive since they operate at mains frequency. Here is a 120VAC type that I use myself to allow safe testing of “live” circuits such as the one you’ve shown. A mains step-down transformer would be a better solution if you want low voltage AC. You could apply a voltage divider to the low-voltage output. Here is a 120VAC wall plug adapter that performs a similar function at a lower mains voltage and with a North American plug.

If we assume that the components shown are locked into a properly insulated enclosure that cannot be accessed without intentional use of tools, then there still remains risk of fire or drawing excessive current that could damage the mains cord. A fuse rated to protect the external wires will mitigate the latter. Resistors such as carbon film can fail in a low resistance mode that does not blow the fuse but produces a lot of heat and smoke.

As mentioned in another answer this has similar safety concerns to that of a “transformerless” capacitive or resistive dropper power supply used some home appliances such as, for example, a kettle where the function keys, heater, display, LEDs etc. cannot come into contact with the user and special “fusible” resistors are used that fail in a predictable and safe way. Such circuits can be made safe enough in such very limited circumstances, but it is far from inherently safe, and generally the rule that we espouse here is that “if you have to ask about transformerless supplies, don’t do it”.

\$\endgroup\$
3
  • \$\begingroup\$ Is this more in reference to end users, and producing a product that is to be consumed by that end user? I.e. something provided as a service to someone who may indeed become harmed if safety protocols (especially locally to that region) are not adhered to. Obviously one must use caution with anything attached to the mains. The immediate concern to my safety still does not seem to be an issue. I will be honest with you, I don't provide products for other people, and to put anyone's mind at rest, I do not intend to (ever). \$\endgroup\$ Commented Jul 24 at 14:44
  • \$\begingroup\$ You are the user if you are using the device. There could also be indirect safety issues (burning down a building, injury due to smoke inhalation, "booby trap" situations with your next of kin or first responders harmed in the process of disconnecting the device post-electrocution) etc. \$\endgroup\$ Commented Jul 24 at 15:15
  • \$\begingroup\$ I will bear that in mind and update my will, and insurance policy. \$\endgroup\$ Commented Jul 24 at 15:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.