0
\$\begingroup\$

I have a WiFi enabled AC mains light dimmer/switch that uses a back to back MOSFET design. The device sits in series with the load (lightbulb) and therefore does not have its own return/neutral connection.

I'd like some help to understand how its MOSFET / transistor arrangement works as I'm not fully confident I do understand how the control circuitry might be managing the MOSFETs and transistors so that:

  • it always has power for itself (a tiny amount of current flows through the lightbulb at all times, but not enough to light it); and

  • so that a controlled amount of current is allowed to flow through it when any level of light brightness is requested.

Ultimately, what I'd like to do is convert the device to be a user input device only (with minimal modifications to the circuit), connecting directly to the supply and neutral - by somehow disconnecting the path between the two side to prevent dead shorts when any level of light brightness is requested.

For example, could I do this simply by disconnecting the base of each transistor so that they're never open and therefore there is never a direct path between the two sides (i.e. pin 3).

This may seem like a pointless exercise/idea, but actually it would then give me a dimmer that I can use in software (over WiFi) as a user input to control other "smart" things such as mood scene selection.

enter image description here

\$\endgroup\$
  • \$\begingroup\$ If the transistor works as variable resistor, like you intend to make a dimmer, it would blow. The dimmer circuits use phase angle control or PWM. \$\endgroup\$ – Marko Buršič Oct 20 '17 at 7:30
  • \$\begingroup\$ Probably not something you should be messing with.. but what are the connections here... esp 1,2,3,4 and 5. maybe add those to the diagram \$\endgroup\$ – Trevor_G Oct 20 '17 at 7:37
  • \$\begingroup\$ @MarkoBuršič it already is a working dimmer, I don't want it to be \$\endgroup\$ – Warren Ashcroft Oct 20 '17 at 8:39
  • \$\begingroup\$ @Trevor they're just the connections to the control circuitry which receives power for itself (5, 1, 2?), and controls the transistor base (3), I suspect (4) is a monitor. The control circuitry is where the WiFi radio and MCU live. \$\endgroup\$ – Warren Ashcroft Oct 20 '17 at 8:41
0
\$\begingroup\$

Use a photovoltaic MOSFET driver like VOM1271 link.

enter image description here

This is a MOSFET solid state relay .

\$\endgroup\$
0
\$\begingroup\$

Disconnecting pin 3 from the control circuitry does the job, this prevents the transistors from opening up and therefore permanently prevents a low resistance path between the two AC sides.

Unless anybody can see any issues with that?

\$\endgroup\$
0
\$\begingroup\$

Your reverse-engineered circuit looks incomplete. I can't make sense of it.

The simplest thing to do would be to work out the minimum load required to keep the device working without modification. You then permanently wire a load resistor of that value into circuit.

If using a resistor rather than a lamp you need to allow adequate cooling - the same as you would for a tungsten light bulb.

This solution will waste a small amount of power. If you add an indicator lamp - a neon, for example - it will act as a reminder to turn the load off.


Another possible solution is to hook it up to a lower voltage supply.

schematic

simulate this circuit – Schematic created using CircuitLab

Figure 1. Low-power setup.

The device probably requires enough voltage and current to power up the internal circuitry. There may be a sweet spot at a lower voltage that will give enough power for the dimmer while minimising the load requirement. Try half-supply?

\$\endgroup\$
  • \$\begingroup\$ I had thought about adding a resistor in series with the mains connection (in place of what would normally be a lightbulb), that way I wouldn't actually need to make any modifications to the circuit at all. However, it need to be something about 1.2k (240v / 200mA) - thats 48w of heat to dissipate, in a small enclosed hole in the wall! \$\endgroup\$ – Warren Ashcroft Oct 20 '17 at 11:12
  • \$\begingroup\$ See the update. \$\endgroup\$ – Transistor Oct 20 '17 at 11:26
0
\$\begingroup\$

I designed several back to back MOSFET dimmers. I am reasonably sure, that: - the schematics has several mistakes. Please double-check your MOSFET gates schematics. - point 5 is a connection to the control CPU's input, sensing the AC to get the Zero Detection - point 1 is a connection to the control CPU's output, controlling each half-sine wave MOSFET switching PWM - point 2 is digital ground

To achieve the MOSFETs to be off, their input must be a low (ground) signal from the control circuitry Before the 1kohm gate resistor pair.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.