A follow up to the previous video on repairing the heater.
A viewer asked how the capacitor diode rectifier gave a 24V output. The key is in the zener regulator, so this vidoe looks at how mains powered zener voltage regulators work, and their limitations.
X class capacitor and self healing.
Zener diode tutorial video: https://www.youtube.com/watch?v=O0ifJ4oVdG4
Dodgy Dangerous Repait video: https://www.youtube.com/watch?v=myqiqUE00fo
Forum: https://www.eevblog.com/forum/blog/eevblog-1482-mains-capacitor-zener-regulator-circuit/
Support the EEVblog on:
Patreon: http://www.patreon.com/eevblog
Odysee: https://odysee.com/ @eevblog:7
Web Site: http://www.eevblog.com
EEVblog2: http://www.youtube.com/EEVblog2
EEVdiscover: https://www.youtube.com/eevdiscover
AliExpress Affiliate: http://s.click.aliexpress.com/e/c2LRpe8g
Buy anything through that link and Dave gets a commission at no cost to you.
Donate With Bitcoin & Other Crypto Currencies!
https://www.eevblog.com/crypto-currency/
T-Shirts: http://teespring.com/stores/eevblog
#ElectronicsCreators #Tutorial #Zener

Hi In a previous video linked up here and down below. if you haven't seen it, I repaired an Arlec Space heater and I did a Dave Cad drawing over basically what was, uh, happening here for the power supply and well, I won't spoil it for you what the failure was, just go and look at it. So all I showed on my Dave Cad drawing was a 47 ohm resistor here, a bridge rectifier, and then the Ac capacitor down here. uh with some bleeder resistors across there and it basically uh, generated the 24 volts output from the bridge rectifier 24 volts uh to power the relay and the uh lead display control, uh, circuitry and stuff for the Uh heater and somebody asked me, well, how do you actually get 24 volts out of the bridge rectifier They didn't quite understand and that's because yeah, I forgot to include the actual regulation side.

So we're actually going to look at a a capacitive drop in Maine's Zener rectifier. Let's check it out. Now, this won't be an in-depth tutorial on Xenodiodes, because I've already done that linked up here, as well as down below. Check that out if you want to work out the intricacies of powering stuff with Zener diodes.

Xenodiodes use a current through them like this to actually, uh, regulate the voltage across them, and when you include the load, the calculations can get a little bit tricky for Zener Diode, so we won't cover that detail here. It's in my previous tutorial video, which is quite extensive on the Zener Diode topic, but basically, um, as you saw on the Pcb, here's a photo of the Pcb. The Zener diodes are actually on the top side. They're through-hole parts, but I've just photoshopped them onto the uh back side of the board so you can see what's happening.

You might think that they look like back to back xenas, but if you actually trace out the circuit, which I've done here, not extensively. It's not a full reverse engineering, but we can actually see what's going on here. because, uh, just the bridge rectifier on its own is not enough to regulate the voltage. You do actually need a regulation element.

All the diode bridge rectifier is doing is current steering. Uh, where the current needs to go? Because I forgot to put it here. you've got 240 volts Ac in we used 240 volts here in Australia, and the bridge rectifier. You've seen you're used to this, uh, configuration of the bridge rectifier, but it's exactly, uh, the same as this.

So let's assume that the Ac waveforms on the positive part of the cycle here. Then we're going to get the diode steering it like that, and we're going to get another diode steering it like that. So then we can get current to flow in our circuit. But when this flips around negative to positive like that, then we're going to get a diode like that and get rid of that one.

And then that's going to jump over there like that and we get rid of that one. and now the current can flow from positive through negative like that, and each alternating uh cycle of the 50 hertz Ac waveform. Then it just gets the current always going in this direction like this never goes backwards and that's the job of the bridge rectifier. But if you've got nothing here, then there's no load.
It actually won't do anything. You'll get no current flowing at all really apart from parasitic capacitances. Unfortunately, I don't have the Pcb with me anymore, so I can't actually measure it, but we were actually getting that 24 volts clamped over there and we saw that, uh, our input capacity. here.

It was an X2 class capacitor, normally 220 nano farads, but it had actually dropped in value to a hundred nanofarads. And this happens to uh, Ac capacitors like this when you get surges on them. The self-healing dielectric inside actually, uh, plasma vaporizes tiny little holes and over every time you get a surge, the dielectric self heals. It doesn't uh, short out.

Usually that's his job. It heals itself by burning the little metal metallized plastic inside the uh poly, put the kettle on capacitor. when I say probably, put the kettle on can be polypropylene polycarbonate. There's various different types of poly material used in capacitors and it's it's metallized on top metal layers.

Then when you get a surge on there, it sort of like punches through a tiny little hole. We're only talking micron size stuff here and then a little plasma arc forms and it just vaporizes the metal in there and then just it forms an insulator again. So you've just got that hole, uh, with the insulator. Here's a little graphic kind of showing what happens there.

and this is why you generally use a self-healing capacitor in a circuit like this. But one of their problems is that over time they can lose capacitance due to, you know, main surges and stuff like that actually causing loss of capacitance due to all these little holes opening up on your capacitor. You've only got so much area on your capacitive plate. Sooner or later, it's going to drop in value.

So we saw a pretty drastic one here. You know, 220 down to 100 nanofarad? But that hundred nanofarads then caused our regulator to drop out of regulation and we weren't getting 24 volts across here. I think we're getting like 12 and a half 13 volts. We're getting roughly half the voltage out of our Zener Diode.

So why does that happen? Well, the effective, uh, resistance of this 220 nanofarad capacitor. It actually goes up in value and the standard capacitive reactive formula Xc is one on 2 Pi fc. The F is the frequency 50 hertz here in Australia C is the capacitance. So if your capacitance value drops, then your resistance goes up.

So that's called the capacitive reactant sometimes. uh, you'll hear people call it impedance and they kind of use interchangeably. But capacitive reactance is actually just the value of the capacitor itself. Impedance also means the capacitive reactance value itself just due to the capacitance plus the internal uh, equivalent series resistance in there.
So like the total resistance, that's what's called impedance. But they're kind of used like interchangeably impedance or reactants. but strictly speaking, they are different terms. So this impedance is what's known as the Ac resistance of the capacitor.

And this is basic Ac circuit theory. The effective resistance of a capacitor is going to change with frequency. But because we've got a fixed frequency 50 hertz, uh, 240 volt mains here, then um, we're always going to get about 14.5 k effective Ac resistance of the capacitor. So basically, that is our dropper resistor for our Zener Diode here because the Zener Diode always needs a dropper resistor in here.

But what is this 47 ohm resistor up here? Well, it's actually because it's very low. It's like 47 ohms compared to 14.5 k Ohms. It's not the dropper resistor for the Zener here, it's actually in rush. uh, protection.

because when you first plug the appliance into the mains, if you've got a capacitor directly across it, which you effectively do with a capacitor, and just assuming your load is like a very low value, then the capacitor appears as a short circuit. So you want to actually have some in-rush protection so that limits the surge current when you first turn it on. And in this particular case it might. I've actually drawn it with a fuse symbol in there because I don't know for certain, but it might be a fusible resistor and that's quite common in this, uh, sort of application.

So if your capacitor does fail a short circuit, then your resistor is going to pop. So this is how you can actually get a power supply directly from the mains. It's a mains driven Zener Diode voltage regulator. But the huge disadvantage of this is that everything in this circuit here is at mains potential.

Okay, it's not isolated. There's no isolation transformer. Even though your circuit might be 24 volts or 3.3 volts to power your circuitry. you don't want to go around touching and probing in your circuit, especially with your oscilloscope done the whole video.

How not to blow up your oscilloscope? You don't want to be connecting your scope ground on here or anything like that. You don't want to be touching it with your wet finger because you really come a gutter. It's a some sort of mains potential. No touchy.

But this sort of Zener regulation circuit directly from the mains is cheap and simple. And that's why it's used in tons. You know you'll find them in uh, light bulbs and all sorts of appliances. Uh, like this because you don't need an isolation transformer and all that sort of stuff that costs a huge amount extra real cheap ass designs.

even do away with the in-rush protection. and you might have like a mob across here as well for like, um, extra surge protection and stuff like that. This particular one didn't So what you've got here is a basic Ac resistor here of 14.5 k stays constant with the 50 Hertz frequency and a Zener diode here. And then you can have your resistor load on there and we can actually work out roughly, uh, how much current.
Uh, we're going to get total, including the Zener current and the load current because you have to separate them. That's in my Zener uh, tutorial and it's roughly 240 volt rms here, divided by 14.5 k. But I know what you're saying, dave. We also have to subtract the Zener voltage here.

Well, in engineering when you're doing ballpark calculations like this, and that's all we're doing in this video is ballpark calculations because 24 volts is like an order of magnitude less than 240 volts If something's like an order of magnitude different to what you're talking about. Like, if you have a resistor that's an order of magnitude higher in parallel with another resistor, you just like ignore it. And that's what we're going to ignore here. Just ignore the Zener voltage.

So, 240 volts divided by 14.5 k. around about 16 milliamps or so through the resistor here and through the zener and or load. But because this load actually is a relay coil and here's the data sheet for it. and you'll see that at a 24 volt relay is actually a 15 milliamp coil current.

So oh, there's not much current left over to actually drive the Zener here, so it's kind of right on the border. And this is why um, when our 220 Nano Farad capacitor dropped in capacitance value and the resistance went up, there just wasn't the current available to maintain regulation in the Zener and the Zener voltage dropped and then we didn't have enough spoiler alert didn't have enough voltage uh to turn on the Uh relay coil. We just didn't have the 24 volts and uh, the actual coil current available because this value here went up in value as the capacitance dropped due to the one over uh formula here. So yeah, and that just starved the relay of current and it couldn't switch on.

So the simplest Ac regulated uh supply that you can get is just basically a capacitor or a resistor. I'll talk about that in a second in series with a Diode bridge and a Zener diode and that's it. Bob's your uncle. You've got a regulated, um, whatever.

24 volts, 3.3 5 volts, whatever it is to power your circuit. But as I said, it's a little bit dangerous. It's not isolated, no touchy, but if you use it inside an enclosed product where the customer can't actually touch it, then meh, it's good enough. So why do they actually use a capacitor? It's got nothing to do with Ac coupling or anything.

You could actually simply use a resistor in here. No worries whatsoever, you could put in a 14.5 k resistor instead of a cap. In fact, it's probably cheaper. Why don't they do that? Well Look at the calculations here.

Okay, so let's calculate the power if we used an actual resistor up here power dissipated in that resistor. So we'll call that Pr and it's I squared. r is the power. uh, formula.
So it's 16 milliamps squared because 240 volts divided by 14.5 k. 16 milliamps squared times 14 and a half k. It's 3.7 watts. Now, that's actually a fairly beefy resistor.

That's a a lot of and it's a lot of power to waste away as well if you're doing a low power circuit or something like that. If you're driving, you know, some Led light or something like that, you don't want to be pissing away 3.7 watts in the resistor. So instead you use a capacitor because there's no power loss. Ideally in an ideal capacitor, there's no power loss at all.

the dissipation, even though its equivalent resistance is 14.5 k. the power loss in this capacitor is effectively zero. or there'll be a tiny amount due to the equivalent series resistance in here. but that's really, really, really small Now, of course, to understand this, you have to get into Power Factor and there's no free lunch here.

Okay, you still have to provide at the generator at the power station somewhere, still has to provide the 3.7 watts. But because this is a mostly capacitive circuit, the power factor is going to be absolutely horrible. So even though the power generator has to deliver the power, the actual load here does not dissipate any power because it's a capacitor. So that's why they whack a capacitor in there.

So um, you know, if especially if it's a little compact device or something, you know you don't want to be wasting. We're only talking like 16 milliamps here. It's not much current. you know I'll be wasting 3.7 watts to deliver your 15 milliamps.

So, but it kind of like puts the problem back onto the grid, but the grid kind of, uh, you know, sort of. They try and balance it out with power factor correction and all the rest of it and we won't go into that. I've probably done a video on that. If I have, I'll link it in.

So this circuit's a bit unusual. I kind of, uh, expected at first glance. um, that they would just, uh, use. You know, have the 24 volt rail and Bob's your uncle.

Maybe a secondary regulator in there to power the 3.3 or 5 volt. uh, digital logic. There's a little micro with a display and stuff like that, but they've actually gone for this configuration here. I haven't drawn the micro in, but basically what it is, they've got a Zener diode up here and a smaller one down here.

So this is, uh, the high voltage one and this capacitor is a 50 volt one. and this is a lower voltage jobby down here. and I believe this lower Zener down here will be like 3.3 or 5 volts whatever is required for the digital logic circuitry. And here's the Uh relay coil here, and the Pmp driver transistor up here.

They've got a back Emf diode across that as well, so that's across the top Zener. but then they're actually uh, rather than then taking that 24 volts and then dropping it down again, they're actually using the return path here for the coil to go through this zener here. So basically the current's going through both and the load is being switched uh, out of this one and then into this one down here. So calculating the power of these two zeners gets a bit complicated depending on whether the relay's on or off, because as I said, the the relay is like, at the relays, nominally like 15 milliamps plus uh, the whatever the lead circuit and your uh, microcontroller takes.
you know, another couple of milliamps there at Leo. Five milliamps or something there, at least, so you know it really doesn't leave anything over for your zener regulation. So this at ballpark calculations seems to be a bit dodgy design, and that's why it was, uh, not tolerant to this value actually? uh, dropping incapacitance when a surges caused that capacitor to self-heal and lose capacitance. And as I showed in my Xena tutorial video, choosing the power rating of your Zener Diode here, you have to take into account your minimum and maximum loads.

In this particular case, depending on whether the heat is on or off, it's going to be like 15 milliamps or whatever difference in the relay current. they might even be running at a lower voltage. You know you don't need exactly 24 volts to operate. that relay is going to have a minimum latch in voltage, minimum latching current.

You'd have to look at the data sheet for that kind of thing. So I think I think because there's hardly any margin in here at all for the extra current. So I think that this top one's not actually 24. It could be, you know, 20 or something like that because we measured 24 total across here.

So then they've got this um, Npn down here and this is actually powered. This is what's powered from the micro. So the micro is powered from like the 3.3 or 5 volts here and that switches on this uh, Npn transistor which then turns on the Pnp. Then you've got a small base current across Uh, both of these Zeners here and then it switches on.

Uh, the relay, which then turns on the heater coil here and we almost forgot about the bleeder resistors across the 220 Nano Farad capacitor here. This is for our safety. So when you pull the thing out, this capacitor could be charged up. You pull it out of the outlet and you don't want to touch the pins because then you can get a zappy.

So uh, you've got two high value resistors 750 k each and uh, so 1.5 meg total across the 220 and that just bleeds the charge off that capacitor. So then the user, um, if they accidentally touch the mains pins, they're not going to get the zap from it. And the reason that they use two resistors, uh, in series physically Here on the board, you can actually see the two There is that, uh, the Smd resistors that they're using. They're only rated for about 200 volts each, so they have to put two in series to get the voltage rating required.
So there you go. I hope you found that. uh, quick follow-up Interesting. As I said, uh, this is just ballpark stuff.

More detail, uh, calculations, and actually have to physically get the board back and do measurements and you know you could go into more detailed stuff. But please watch that Zener tutorial if you want to know all about uh, using Zener Diode circuits for regulation. But that's what they're doing here. It's basically a capacitive.

It's this is not what's called a capacitive divider power supply, which is basically uh, two capacitors, um, two or more capacitors in circuit and then using that Ac resistance to actually tap off a smaller voltage than the 240 volt you're feeding in there. Basically, this is just a zener circuit here, which as I said, you could use a resistor here, or you could use a capacitor. They use a capacitor because they want to get the power dissipation down and then, um, voice the problem off onto the power generator. So if you found that video interesting, please give it a big thumbs up.

And as always, discuss down below and check me out on my alternative platforms: Odyssey and Utreon as well. I've got like 65 000 subscribers over on Odyssey, as well as some exclusive content over there as well. Catch you next time.

Avatar photo

By YTB

25 thoughts on “Eevblog 1482 – mains capacitor zener regulator circuit”
  1. Avataaar/Circle Created with python_avatars Derek Loudon says:

    Hi Dave, these brief 'explanation' videos are a very useful resource for 'faded memory' refreshers (I'm a long retired engineer). Thank you ☺️

  2. Avataaar/Circle Created with python_avatars Shane Van Ingen says:

    Enjoy these whiteboard videos.
    Though you weren't able to explain the whole circuit without spoilers 🤣

  3. Avataaar/Circle Created with python_avatars Rx7man says:

    Considering this device is actually called a HEATER, some 3.7 watt power dissipation from a resistive dropper wouldn't worry me too much.. if it was in a light or something else then I'd be more concerned

  4. Avataaar/Circle Created with python_avatars Oliver Schönrock says:

    Small correction. The generator has to provide 3.7VA not 3.7W.. Or perhaps just "the generator has provide the 'current', ie the 16mA, but this is 90 degrees out of phase, so it amounts to 'no power'" And actually quite often the "generator" doesn't have to provide that extra current at all, because there will be an inductive load somewhere more locally which just supplies (and hence offsets) the capacitive "leading" current.

  5. Avataaar/Circle Created with python_avatars Control Man says:

    Fun video dave, this and the repair video. And quite informative ….

  6. Avataaar/Circle Created with python_avatars Johnson Lam says:

    Thank you very much, another fun learning video.

  7. Avataaar/Circle Created with python_avatars James John says:

    good explanation thanks

  8. Avataaar/Circle Created with python_avatars Topher Teardowns says:

    Is there a particular reason they use 24v relays instead of 12v? If they are pinching pennies, wouldnt it save them twice over. (Able to use a lower value cap), etc..

  9. Avataaar/Circle Created with python_avatars veganath says:

    Wow! That I found this in it's entirety intriguing speaks volumes of you terrific delivery, thanks Dave

  10. Avataaar/Circle Created with python_avatars onjoFilms says:

    Interesting stuff.

  11. Avataaar/Circle Created with python_avatars Terry Smithwick says:

    Plugging in a capacitive load into the electrical grid is not a problem. Yes, there is (reactive?) power that the grid needs to supply and then accept back each cycle; this produces non-imaginary current flowing through wires that heats them up and has to be made up for by the generators.
      But this is capacitive reactance, and is 180° in phase from inductive reactance. The (US) national grid operates on a power factor of 0.8, which is inductive. Most of the loads on the grid are inductive; transformers and motors mostly. That means the generators have to supply power to build up the magnetic field in all this inductance, which heats wires &etc.
      I once worked for a company with many large assembly lines. They had many large motors each and we used heaters with some inductance themselves. In order to prevent extra fees because we were inductive-loading the power factor out of tolerance, we had a Capacitor Farm. All those magnetic fields growing and shrinking in our motors were feeding power into our own capacitors locally, so the generators saw a normal load and the long-distance wires to the power company could stay cool.
      So if you plug a capacitive load into your house receptacle, rejoice in knowing that you're just balancing out a bit of your refrigerator's inductive load, and making those generators a bit less loaded.

  12. Avataaar/Circle Created with python_avatars TheCod3r says:

    Love these whiteboard videos Dave. By far the best series for me

  13. Avataaar/Circle Created with python_avatars Seventh Anubis says:

    If the stupidity of it is to save costs, why didn't they just use a bimetalic thermostat? For the relay, LEDs, and electronics, may as well use a switch mode power supply.

  14. Avataaar/Circle Created with python_avatars Gus Martin says:

    If you take the capacitor apart and unroll the layers, I wonder if you could see the damage? What would that look like under the microscope?

  15. Avataaar/Circle Created with python_avatars φ says:

    Polyputaketalon, 👌🤣👍 ❗

  16. Avataaar/Circle Created with python_avatars TheHue's SciTech says:

    Worth calling out that the strange-looking stacking of 24V and 3V3 kinda allows the microcontroller to be powered "for free"-ish in some vague sense. I.e., a 240V AC mains + dropper cap = a high compliance, practically-fixed-current AC current source; if you were to run the MCU off a LDO off the 24V, then the current for the MCU and the current for the relay coil would be added up. But with those elements in series, it's only the compliance voltage of the current source that needs to be increased, and since it's already at 240V, it's all good to go.

  17. Avataaar/Circle Created with python_avatars Keri Szafir says:

    The Zener regulator makes perfect sense with the series resistance impedance on the AC side and low current loads.
    Good that you give the safety warnings, they're absolutely essential with this type of circuits.

    BTW it looks like I'm starting to teach people electronics.
    Ttoday at our local hackerspace I had to explain the FULL BRIDGE RECTIFIER's principle of operation to one guy who has a university degree in IT but was not taught the basics of EE.

    He had a pair of IN-12B Nixie tubes that he wanted to use in a project, only didn't know how to wire them… poor thing. I breadboarded a simple 200VDC PSU with two power transformers connected back to back for galvanic separation from mains, and taught him how to control these nixies using transistors.

  18. Avataaar/Circle Created with python_avatars Name Redacted says:

    Anybody who tries, year after year by different people in different countries, to use power supplies without galvanic isolation, should be de-balled. If such a person (male implied) survives the operation, then this same person should be shot. If they should survive being shot, they shot be shot again!!! Even my 1980's textbook warned me against this circuit, but I actually built this circuit, only to shock myself (only mentally, not physically), when I measured unloaded output voltage!!!

  19. Avataaar/Circle Created with python_avatars Scott Pelletier says:

    Very cool. I had an LED bulb go out so I smashed the power supply out of it to see how it worked. This pretty much looks exactly like what I saw… Simple rectifier spitting 50 to 60 volts DC out to run a panel of 50 little LEDs. Great explainer!

  20. Avataaar/Circle Created with python_avatars atkelar says:

    I came across a resistor based version of this in the standby circuit of the Revox power amp restoration I recently completed. It gets way too toasty for my taste, so I disconnected it; besides, the optocoupler that turns on the mains is probably shot too. I replaced the original with a higher value to cut down the current, but it still was too hot even after short periods of powering it. Note: I'm convinced the zener would have properly conducted, since the original value was also supposed to work across 110V and I have 240V here, so double R should not be an issue for the current…

  21. Avataaar/Circle Created with python_avatars keith king says:

    Nice stuff Thanks Dave 😊 👍

  22. Avataaar/Circle Created with python_avatars Don Matejek says:

    Great explanation, Dave about the loss of capacitance, not creating enough voltage, which caused the problems further down the circuit. Hope that made sense.

  23. Avataaar/Circle Created with python_avatars Don Matejek says:

    "Come a Gutsa". Gotta love Dave's dielect! Lol.

  24. Avataaar/Circle Created with python_avatars stelmo502 says:

    ALWAYS wanted to know about these type of circuts. THANKS

  25. Avataaar/Circle Created with python_avatars Fred Flintstone says:

    wouldn't you have 240 x 1.4 so really 330 volts after the bridge???

Leave a Reply

Your email address will not be published. Required fields are marked *