Whether you should actually use a generic Chinese touch switch is debatable. They don't seem to fully understand electrical isolation much. A version sold by a prominent local supplier is more likely to be built to a safer standard. This module came with four wire nuts, which are rarely used in the UK.
The bulk of the work on the PCB is being done by a dedicated chip that is not the typical microcontroller. It has dedicated circuitry including an amplifier and filter for the touch input and timing for the triac driver. The use of a capacitor in series with the triac gate is to keep the power supply simpler, as it results in a short trigger pulse of reasonable current.
The data sheet for the chip has extra layers of safety on the touch input, both for the user and the chip. The extra components have been omitted in this design. The timing resistor is an odd value. Possibly a "safe" value for both 50 and 60Hz that ensures the triac trigger points are well within each half sinewave.
The triac is very small. Its rating is a surprisingly high 600mA which suggests a typical lamp rating of 60W or much lower with LED lamps due to the high inrush current associated with their circuitry. The most common failure with dimmer modules like this is the triac going short circuit when a lamp fails with a pulse of high current.
I specifically avoid using intrusive mid-run video adverts for a more relaxing viewing experience.
If you enjoy these videos you can help support the channel with a dollar or two for coffee, cookies and random gadgets for disassembly at:-
https://www.patreon.com/bigclive
Patreon supporters get early access to advert-free videos as they are made, and also regular live streams.
The Patreon support keeps the channel independent of YouTube's advertising algorithms allowing it to be a bit more dangerous and naughty.
Other contribution options are available at:-
http://www.bigclive.com/coffee.htm
#ElectronicsCreators

Let's take a look at one of the many touch switch controllers on eBay and to be honest, I don't really recommend buying it from eBay because it, well, there are no standards on eBay You get random stuff, but the idea is with this controller, you've got four wires. You've got the brown Live Wire you've got the blue neutral wire which is common with the lights and you've got the red switched wire that goes out to the lights and you've got this black wire that comes out. The colors may be different. This is one of these things that I you know I've told you the colors on this one, but don't uh, just assume they're going to be the same with other controllers. Who knows the color codes in some of them? But this one with the ring crimp in it is the touch sensor and it's labeled on the back just as shown connected to a ring crimp. There's actually, if not marked, live, neutral or anything in this. However, I've rigged this up to show you a tolerating to some Christmas lights tungsten Christmas lights because they just seem an appropriate load and this is better for Tungsten loads. It is usable with LED lights, but uh, it is really just probably better suited with Facebook dimming, which it does to tungsten, noting that there's no interference suppression. So here's a little ring crimp and I've just touched the wire and it triggered it. So when you touch it, it comes on the low setting. Touch it again. medium Sen touch it. Oh that's wrong. Uh, low, medium high off low medium high I'm not sure that is full brightness, but that is probably part of that's configurable in the circuitry so to speak. This is hackable in a way and you can make it safer by putting a big resistor. Oh look, it even responds to just touching that that's very sensitive into just touching the insulation. but you could put a resistor in line with this if you wanted extra safety. This requires the use of a either a metal touch pad or a metal case light fitting that isn't grounded. It can't be grounded, it needs to use it as the antenna to pick the signal up. So looking inside this little box Baja it is clipped together by two clips on the other side, one of which I've already broken off because the first thing I did when I got this was I looked at it to see does it have any proper electrical separation? The unit is unplugged I feel they need to point that out so I can show you this is good surface mount in the back. it's got some physical components in the front. one moment please. I'm just going to get the uh this photographed and then we can continue. The reverse engineering is complete. Let's Zoom down a little bit in this and take a look at the circuitry. So this is the through-hole side of the circuit board. I've flipped the image so it correlates to what's down below. So this resistor here is part of the power supply blue, gray, orange, 683, six to eight thousand ohms that goes to a diode for half wave rectification and then to this smoothing capacitor, which is about 47 microfarad at 16 volt. and there's also a Zener or Zener diode on the other side the circuit board that caps the voltage to around about 6.8 volts. There is a track here, and a space for a bigger track, noting that this circuit provides quite low gate current just to keep the circuitry simple in the purse Supply symbol And it's only in pulses, so you'd have to use a very sensitive Triad but there is the option there to upgrade it if desired. This capacitor here is part of a filter circuit for the touch input, and this capacitor here is between you and domains. It's supposed to be two capacitors, but it has one and it is a 3kv Um 470 picofarad capacitor. So let you take a picture of that if you want to play along here and then I shall show you the other side of the circuit board. Things are worth mentioning in the earlier part of the video I Said it has to be used under sort of ungrounded metal frame, but light. Don't remove the ground if there's one on it. if that's the case. Uh, it's kind of designed for double insulated lights with a metal housing. If that's not the case, then you can theoretically put a rubber grommet and had a little touch pad onto the front of it using this circuit, even if the case is grounded and just touch that pad instead. On the other side of the circuit board, we have the rest of the power supply circuitry. So here's the Hat resistor on the other side, through the diode to the smoothing capacitor. Um, and there's this Zinner or Zener diode across that and that provides power to the chip. It's very low current. It's interesting to note that that case is labeled 110 to 240 volt. They've gone for a universal approach here, and they've compromised with some component values to try and fit that in. This resistor is possibly disputing a little bit more power, but having said that, it's a one watt resistable look of it and it's only dissipating about 0.4 Watts Um, Things worthy of Newton here. Um, the filter circuit is based around these capacitors. I'll show you that in the schematic, there's the triac drive circuitry which is a very interest because as a capacitor in Sears with the gate. Um, yes, this resistor to set the clock frequency. That's something you could play with if you wanted. It will vary the sort of ratio of brightness in the multiple settings, and this resistor isn't really worth changing. This is the one that's used to get the reference so it knows where the start of the sine wave is. Let me bring in the schematic. Here is the schematic: I shall Zoom down just a tiny bit more just to make that nice and big. So here's a power supply. Note that Live is labeled zero volts here. That's because all the circuitry is reference saliva. That's for two reasons. It is uh, switching the light to live via the triac, but also because it's relying on when you touch the pad. it's looking for leakage from the main side which the circuit is referenced to to ground via the capacitor and you touching it which provides the link and it needs to have its zero volt reference to Live. Uh, if you have one of these and it's not working very well, check the polarity of your wiring is correct and that uh, the Live is definitely live because otherwise it may not work. So the power supply is that resistor that diode charging that capacitor with the Zener Diode or Zener diode across it. If you prefer to shunt the voltage level to about 6.8 volts. I Measured 7 volts. the touch input has the pad. It's got the 470 picofard. The data sheet shows to a 1000 picofard or one nanofarad capacitors in series at rated one kilovolts each. They've used a the equivalent capacitance value, but with a single three kilovolt capacitor. I Just prefer. You know, in situations like this, you're better using multiple components for extra separation. they've also got a 3K resistor. The data sheet shows a 1K there and also a diode to the positive and negative reels, possibly to allow for the touching of that with a static discharge and the diodes would help divert it. While that limited the current, in the case of the chip itself, it may have the diodes inside and maybe that's why they used a higher value there. Um, the triac is driven via this 47 nanofarad capacitor. a 470 Ohm resistor is that 471? Let me just grab that. Yes, it is. I've double checked 471, 470 4710 and the triac, which is a Mac 9786 has a 10K resistor pulling it to the zero volt rail live to keep it turned off. and when this circuit sends a positive going pulse to that track, this capacitor limits how much current can flow. Switch: Just a single pulse is used to trigger that track. It doesn't need a continuous drive, all it needs is a pulse to turn it on and it will latch on. I'll show you the sine wave data afterwards so the chip knows where to get its timing from. in the sine wave, it has a resistor. a high value: 1.5 Meg ohm. Two resistors would have been nicer and serious as to the voltage rating, but it's connected to neutral, so it can tell when the sine wave has gone past the zero. Crossing Point That's when it changes priority and is effectively at the start of each half. There's a resistor here, which is not the value shown on the data sheet. it's much higher. This controls the oscillator speed inside for the timing. from the pointer to text the zero crossing point to the point it triggers the triac and the sine wave that could be experimented with. You could use our base value of resistor and a potential from that tweaker and that will vary the intensities associated with each sort of level. Um, as it is the circuitry it I Don't think it has a I Don't think it covers the Fool's sine wave. It's the lamp is never really going to reach full intensity. I Don't think? Not sure about that. Um. Or they might have just scaled up into the sine wave to make it so that it doesn't go too dim. That's possible as well. Uh, but that's good if it doesn't trigger too early in the sine wave, because, uh, that results in, uh, longer lamp life. This little pair of components here is our capacitor and a discharge resistor across it presumably that looks as though it's associated with the touch input. and it may be to add a slight time delay so that it's less prone to glitches and spikes that could false trigger it. and it requires a decisive touch. Hotels? Oh, there is no suppression circuitry. There's no inductor. The radio hams are not going to be too pleased about that because this means that this thing will cause electrical noise on the EM audio Spectrum radio spectrum. Anything else here? That is it. You could play around with these values. You could play around with that value. Um, that's more or less it. They're the two main things. You could experiment with. this capacitor value here in this uh, and it's associated with resistor and this one for the timing, which will change the sort of intensities. the waveform. The zero crossing point of a sine wave is when it goes from one priority to the other and it passes the zero volt line. Two things happen there: If the track was on. It will turn off because there's no current flowing through and it's a latching device, so it turns off at that point, ready for triggering the next half. It also starts a timer that is based on the value of that resistor and the number of times you touched it. To say, say, for instance, half intensity. It will time it until it's roughly halfway along the same wave. It will fire the track. The track will latch on, and the lamp will light for that half of the sine wave. Then when it crosses over, it starts the timer again, but it then fires and the other half of the sine wave for much lower intensity. It might actually be a much longer time delay and uh, fire at the end of the sine wave and the same again for the other half. For full intensity, will basically fire early on. So I wonder if that resistor will really affect the the full intensity then, because I'd expect for almost as soon as it detected the zero Crossing Point A very short delay afterwards, but still probably worth experimenting with, but that is more or less it. There's another possibility here. Did you see how it triggered when I touched the wire? You could theoretically put a metal pad in the back of a plastic housing and then touch the front of it. It may actually detect the capacitance through the plastic onto your fingers, but it's interesting. The data. Oh, that's wrong. The data sheet for the Tt6061a is available online. You can find that in Google It's an interesting circuit, but as I say, use with caution of buying from from eBay That's one where that thing I suggested they're putting a bit of foil behind the plastic case and detecting capacity through. That would actually be maybe a safer option. It's probably worth experimenting with. That and another thing you could do: the black lead that goes out. for the touch sensor. You could put a resistor, another resistor, or another capacitor in series of that just to act as extra insulation between you and that, but that is it. It's an interesting little circuit, and it does seem to work, so there's not much else to say about it. There are other versions. You may be able to find a version that when you touch it it either turns straight on, straight off or as you press the pad it will gradually increase intensity to fix level or when you touch it I Get take your finger off and touch and hold it again. It reduces intensity which is kind of more versatile. but this one low medium high is very simple. It's very straightforward, but that is it. A very simple circuit thanks to this chip and well it works.

16 thoughts on “Ebay mains voltage touch dimmer – with schematic”
  1. Avataaar/Circle Created with python_avatars Alexander Grรคf says:

    Interestingly, I replaced exactly such a device in a touch lamp with one that is a simple switch, because all the smarts has moved into the bulbs. And they're having a hard time with dimmers.

  2. Avataaar/Circle Created with python_avatars colin smith says:

    As a radio ham,I have a touch lamp and it does generate a large quantity of noise,also when I am transmitting the lamp turns on and of because the touch wire act as an Antenna,no amount of suppression seems to work.

  3. Avataaar/Circle Created with python_avatars John Nodge says:

    I got given a faulty touch light by a neighbour to look at.
    It had that exact module inside, the touch ring was under the upstand threaded nut which clamped it to the base. The whole light was made of metal, and it wasn't earthed, so I decided it was too dangerous to repair and return to them, so they got it back faulty.

  4. Avataaar/Circle Created with python_avatars ะ ัƒััะบะธะน ะฑะพั‚ says:

    Says "shit" and picks himself up on it, and then proceeds to say titty and totally let's it slide.

  5. Avataaar/Circle Created with python_avatars Hola! Cucumber Man says:

    Looks very hackable and useful thanks for bringing this to our attention Clive ๐Ÿ˜Š

  6. Avataaar/Circle Created with python_avatars Electronic Passion says:

    ๐Ÿ‘๐Ÿ‘๐Ÿ‘‹๐Ÿ‘‹

  7. Avataaar/Circle Created with python_avatars Brian Terrill says:

    So… it uses you as capacitance grounding to change the circuit??? That can be scary in wet winter conditions and a bad circuit.

  8. Avataaar/Circle Created with python_avatars SnekliwFilms says:

    Ebay, mains and touch! Sounds like a good time!

  9. Avataaar/Circle Created with python_avatars twocvbloke says:

    Reminds me of when Alec of Technology Connections made his go boom due to ambiguous wire colours, and his was sold in a store, dodgy stuff… ๐Ÿ˜›

  10. Avataaar/Circle Created with python_avatars imark7777777 says:

    Well I'm trying to procrastinate hey what's that over there is it a squirrel? This reminds me of an experience. I was visiting the Philippines 220-240v 60 Hz and the place I was staying the person had a US 120V AM/FM cd player radio thing with touch controls. This was plugged into a step down transformer. There was one issue one was required to time jumping up off the ground and touching the buttons for the buttons to register the touch input.
    I suspect this was due to capacitance?

    That is until a day later from being shown this, I was accused of plugging it in with out the transformer. yeah the guy who brought a multi meter wouldn't check the voltage before plugging something in in a foreign country after being shown it plugged in and requiring a transformer yeahโ€ฆ Somebody else in the group who never admitted it. Slightly off topic what was I talking about oh yeah the required conductive gap between the ground the physical ground and the human operating the buttons made me think about all the times I've had issues with touch controls and how annoying they are.

  11. Avataaar/Circle Created with python_avatars Ralf B says:

    So this one would not work reliably in countries with Schuko or Euro receptacles?

  12. Avataaar/Circle Created with python_avatars t3hbeowulf says:

    "The Hams are not going to like this…"
    That's okay, 10-15 minutes of transmitting at 50-100W will blow this module up anyway, noise gone. It's a self-correcting problem.

  13. Avataaar/Circle Created with python_avatars imark7777777 says:

    A long while ago I found a fancy cheapish touch dimming lamp on the side of the road in somebody's trash.
    I decided to disassemble it and i took the module out and stuck it into a plastic outlet box and attached the Touch line to the ground of a three prong outlet and a phone jack and crock cable. Wired up with a two prong cord and metal face plate. This let me touch the plate to turn the light on or crock cable that went to a strand of cat cable with a loop on the end which wrapped around the cord of a desk clip lamp that I had clipped to a bookcase. This let me tuck most of the wires out of The Way and I could touch a wire by the clip to activate or the face plate or I could plug a phone cable in with a little break out to route a wire around somewhere more convenient.
    After like 20 years the module finally died so I replaced it and it was far cheaper than I expected to just go down to the local hardware store. The best part is the outlet cover that I used was custom painted by my mom with a little Bugs Bunny figure on it and that's where it has lived since.

    Considering that most of the lamps in the US are two prong whether it's metal or plastic double insulated or notโ€ฆ Three prong ones supposedly exist I have never seen one in the wild.
    other than cable abrasion the only contact point is in the socket connections which is usually insulated with a cardboard insert that if in good condition is sufficient, and I have never scene one otherwise unless the whole lamp is soaked which you wouldn't want to use anyway at that point.
    I have not had any issues and I find this fairly safe knowing what it is and how it works. although now I'm thinking what's three prong metal and motorized that I could plug in in touch to control speed hmmmm.

  14. Avataaar/Circle Created with python_avatars DJ kurushimeru says:

    t h e d a t a s h i t oh thats not right

  15. Avataaar/Circle Created with python_avatars Janus Kobain says:

    I totally wouldn't trust my life to how good some random manufacturer was at isolating the touch pad, lol.

  16. Avataaar/Circle Created with python_avatars Colin Burke says:

    Hey hey there, Colin Burke, Lead/Principal EE, embedded engineer, MatE for Swift Sensors. We make unique long-range, ultra low power Bluetooth sensors.

    would you ever want to reverse engineer one of our low powered sensors? Is there a formal contact for you which we could discuss this further?

    While it's true there are hundreds of sensor companies out there- I do believe our products are unique and incredible quality. Of course I am biased, but I have never seen a Better product made by just ~2 people in 10 months including all electrical, mechanical, emebedded, and MFG work.

    We would love to have you review our circuitry and function- we have all different kinds of sensors from Temp, vibration, to Current Transformers (up to 200Arms), VACrms (0-500), and even some very unique things like a matchbox size TWO channel thermocouple sensor- which supports standard TC mini plugs and all TC varieties (K, J, T, B, E, S, more) that lasts 5yrs on 2xAAA taking measurements each minute.

    Many of our other sensors last 10+yrs. Anywho! Let me know if you'd be interested or if there is a formal contact and we will be in touch!!

Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.