Generating a DC supply from the attached circuit

[burtoninlondon]

I have a query...

Is it possible to generate a stable 5V @ 100mA DC voltage from the attached circuit without affecting its primary functionality?

 

[AnalogKid]

Yes, but the the 12 12K resistors will have to change and will get very hot.

This circuit has several glaring errors. For example, when Q1 is turned on it shorts the AC LINE directly to GND.

Post a larger schematic that is readable. Also, a link to wherever you got this one.

Note that using power MOSFETs in a lamp dimmer circuit was patented back in the 90's with a circuit way less complicated.

ak

 

[burtoninlondon]

AnalogKid, thanks for the reply.

The circuit has been taken from an application note for a development board designed by ST Microelectronics.

A clearer schematic and application note is available here...(http://www.st.com/content/st_com/en...olution-eval-boards/steval-ild005v1.html)...I would have uploaded the application note but the file is too large for this forum.

Note that the schematic I posted had an error, the lamp is not connected directly to the circuit, the lamp is connected in-line with a power supply with the neutral for the circuit being made available through the lamp.

Can you share more about your comment on the simpler lamp dimmer circuit using MOSFETs?

 

[chopnhack]

A clearer schematic and application note is available here...(http://www.st.com/content/st_com/en...olution-eval-boards/steval-ild005v1.html)...I would have uploaded the application note but the file is too large for this forum.

I get page not found when I click the link.

 

[burtoninlondon]

Try this

http://www.st.com/content/st_com/en...ing-solution-eval-boards/steval-ild005v1.html

 

[chopnhack]

http://www.st.com/content/ccc/resou...df/jcr:content/translations/en.DM00161788.pdf
The schematic link on their website seems to be down, but the above link is for the datasheet which shows the schematic on page two.
Is the circuit something already in use that you want to scab on to?

 

[AnalogKid]

That is not your normal dimmer circuit, and the 48-page description is excellent. I was going to point out the lack of a decoupling cap on the CD4025, but now it makes sense.

Other circuits are in US patent 4649302 (from 1987), 5004969, and 7102902. Note that many FET-based dimmer circuits have a full wave bridge so you need only one power MOSFET, but to me this defeats the better efficiency of the MOSFET (over a TRIAC) by adding the power loss and heat of the bridge.

If you want a smarter controller, consider the chips from LSI/CSI: http://www.lsicsi.com

ak

 

[burtoninlondon]

http://www.st.com/content/ccc/resou...df/jcr:content/translations/en.DM00161788.pdf
The schematic link on their website seems to be down, but the above link is for the datasheet which shows the schematic on page two.
Is the circuit something already in use that you want to scab on to?

Yes, I bought this development board for a trial and thought the dimming performance was acceptable.

I want to add a secondary digital circuit to this one, hence the query of generating a DC voltage.

As a bit of extra information this dimmer circuit is perfect for use in the UK where typical wiring installations do not have a neutral conductor available at the point of installation (the light back box). I understand this is not the same in other parts of the world where it is more common for a neutral to be present.

 

[chopnhack]

As a bit of extra information this dimmer circuit is perfect for use in the UK where typical wiring installations do not have a neutral conductor available at the point of installation (the light back box)

What do you have coming into the box? Power has to come in on one wire and leave on another, the names may be different, but physics are the same. I thought UK has red (live) and green/yellow earth as well as a blue neutral?

What I was going to mention was a small transformer and rectifier setup if you had the space

 

[burtoninlondon]

What do you have coming into the box? Power has to come in on one wire and leave on another, the names may be different, but physics are the same. I thought UK has red (live) and green/yellow earth as well as a blue neutral?

What I was going to mention was a small transformer and rectifier setup if you had the space

Typically at the light switch in the wall there will be an unswitched live conductor and a switched live conductor, these both connect into the light switch (this is called 2-wire configuration which is most common in the UK). These would then be routed to the lamp in the ceiling where the neutral is available therefore completing the circuit. It is common for an earth to be present at the light switch but this cannot be used as the circuits return path.

The circuit shown used a zener diode to clamp the voltage to 15V at the output side of the rectifier circuit, is it too simplistic to suggest a second zener diode in parallel to clamp the sub-circuits voltage to a lower level?

 

[chopnhack]

Typically at the light switch in the wall there will be an unswitched live conductor and a switched live conductor, these both connect into the light switch (this is called 2-wire configuration which is most common in the UK). These would then be routed to the lamp in the ceiling where the neutral is available therefore completing the circuit. It is common for an earth to be present at the light switch but this cannot be used as the circuits return path.

Pardon my ignorance, I will have to look into uk wiring more - I forgot you have 230V. In the states, we have split phase, we use two 120 legs to derive our 240v and in our circuits for 240, its two black conductors each are live with 120v and one ground earth. Each 120 leg crosses over the other wire for return. I would assume the same happens in your setup so I am not sure why the neutral would be needed.

is it too simplistic to suggest a second zener diode in parallel to clamp the sub-circuits voltage to a lower level?

Diodes never behave ideally so if you have two in parallel, chances are one of them will have a lower forward voltage and will end up taking all the current.

 

[AnalogKid]

Typically at the light switch in the wall there will be an unswitched live conductor and a switched live conductor, these both connect into the light switch (this is called 2-wire configuration which is most common in the UK). These would then be routed to the lamp in the ceiling where the neutral is available therefore completing the circuit.

About the same as in the US. Usually we have unswitched line and neutral coming into the switch box, and switched line and unswitched neutral going out to the light. The two neutrals are joined together inside the box with a wire nut.

The circuit shown used a zener diode to clamp the voltage to 15V at the output side of the rectifier circuit, is it too simplistic to suggest a second zener diode in parallel to clamp the sub-circuits voltage to a lower level?

Yes. Ohm's Law. With the AC peak voltage and the equivalent of all of those 12 K resistors, you can calculate the peak current available at the zener for both the dimmer circuit and whatever else you want to run. Hint - the guys at ST probably have designed the circuit for exactly what the circuit needs at low line voltage.

ak

 

[burtoninlondon]

With the AC peak voltage and the equivalent of all of those 12 K resistors, you can calculate the peak current available at the zener for both the dimmer circuit and whatever else you want to run. Hint - the guys at ST probably have designed the circuit for exactly what the circuit needs at low line voltage.

Ok, so in simplistic terms would it feasible to replace the existing 15V zener diode with a 5V zener diode then resize the 12K resistors to draw the appropriate current for the dimmer circuit and the 100mA I need? This obviously assumes that the dimmer circuit will work at the reduced voltage.

Or is there another way?

 

[Bluejets]

One could get hung by the authorities if this "generated" low voltage was brought out into the real world without approved isolation.

 

[burtoninlondon]

One could get hung by the authorities if this "generated" low voltage was brought out into the real world without approved isolation.

Understood...but be assured that this is a theoretical query and I have no intention of cobbling a solution together myself.

I'm simply assessing the feasibility before I engage with a professional who will ensure compliance. If there is no feasible solution then the idea is shelved.

 

[davenn]

I suggest you work directly with a pro tech from this point onwards

this forum ( and most others) really don't approve of giving advice to direct mains powered circuits like this

the general view is that if you have to ask these questions, then you probably really don't have the experience with such and should be playing with circuits like this

and for that reason, I will close the thread


Dave