Need design advice:

[aruna1]

Hi All,
this is an extension to my thread

Transformerless power supply: Inrush current limiting resistor placement

Hi All, I'm studying transformerless power supply systems these days and noticed that basically in every circuit I found on internet has a inrush current limiting resistor connected to the supply line where capacitor is not connected. That is, If the dropper capacitor (and its parallel...

www.electro-tech-online.com

I'm trying to build a over voltage/ under voltage detection circuit for 230V AC mains

for this I need
small power supply for the circuit
way to measure AC voltage

for the power supply, I'm planning to go with transformerless power supply
to measure, I need to drop AC mains to a smaller value, and then use a window comparator.

now since I'm not using a transformer, I've been scratching my head for few days now on how to make ground common for above requirement

Currently I'm thinking doing something like below [see how I connected grounds of two bridge rectifiers?]

logically, I feel like the circuit makes sense, but I also have a gut feeling that this will not work.

Please share your advice on this. [I know I should avoid TPS, but I don't have enough space to use a transformer, in the enclosure]

Thank you very much

 

[DerStrom8]

I cannot recommend strongly enough AGAINST using a transformerless supply. They are incredibly dangerous as there is no isolation from mains, and one small fault in your circuit could lead to a very deadly situation. Please please PLEASE reconsider your decision to not go with a transformer-based supply.

Best,
Matt

 

[gophert]

You mean...

Please please PLEASE reconsider your decision to not go with a transformer-LESS supply.

 

[aruna1]

I cannot recommend strongly enough AGAINST using a transformerless supply. They are incredibly dangerous as there is no isolation from mains, and one small fault in your circuit could lead to a very deadly situation. Please please PLEASE reconsider your decision to not go with a transformer-based supply.

Best,
Matt

Hi Matt,
Yes I'm aware of the danger. but for the sake of design, what is your opinion on above?

 

[DerStrom8]

You mean...

No, re-read it.

 

[gophert]

No, re-read it.

Right, right. Never mind.

 

[DerStrom8]

Hi Matt,
Yes I'm aware of the danger. but for the sake of design, what is your opinion on above?

I stated my opinion already, and providing further answers to help you actually create this thing will only encourage careless design practices. It's not worth it, in my opinion.

 

[JimB]

[see how I connected grounds of two bridge rectifiers?]

Yes, and if you trace the current paths, on the negative half cycles of the mains you are bypassing your current limiting 1uF capacitor.
Something will probably go bang.

What does this thing connect to?
Unless it is completely isolated so that it is not possible to touch any conductive part during normal operation, this is a SERIOUSLY BAD IDEA.

JimB

 

[Visitor]

Microchip recommends a capacitive dropper in their power meter circuits, where there is ABSOLUTELY NO CHANCE of a human coming into contact with ANY PART of the circuit. They are very explicit about this.

 

[aruna1]

Yes, and if you trace the current paths, on the negative half cycles of the mains you are bypassing your current limiting 1uF capacitor.
Something will probably go bang.

What does this thing connect to?
Unless it is completely isolated so that it is not possible to touch any conductive part during normal operation, this is a SERIOUSLY BAD IDEA.

JimB

Hi Jim,

circuit will be used in a fully enclosed system with no external access. As I mentioned this for a over voltage under voltage detector

 

[AnalogKid]

If you rearrange the circuit blocks, there will be only one bridge rectifier needed and both the 12 V DC output and the scaled-down rectified AC output will have a common GND.

Move R1-R2-C1 to the other side of the bridge, between the bridge + output and D1.

Put the R3-R4 divider between the bridge + output and R2.

NOTE: Update your will before trying this.

ak

 

[aruna1]

If you rearrange the circuit blocks, there will be only one bridge rectifier needed and both the 12 V DC output and the scaled-down rectified AC output will have a common GND.

Move R1-R2-C1 to the other side of the bridge, between the bridge + output and D1.

Put the R3-R4 divider between the bridge + output and R2.

NOTE: Update your will before trying this.

ak

Hi Ak,
I believe moving C1 to other side will not work as to to use capacitor reactance we need AC, and other side of bridge is DC

 

[Diver300]

If you've got space, you can get some quite small mains transformers, that would be reasonably linear with voltage as well as providing the power.

If you are going to use a transformerless supply, I suggest using a shaver transformer to test with. That will reduce the likelyhood of getting a shock.

To measure the current with a transformerless supply, you can simply measure the current through the limiting capacitor and use just one bridge rectifier. Potential dividers can use capacitors when on ac. That is how x10 oscilloscope leads work. If you use a capacitor / resistor potential divider, it will rely on the frequency being correct, but that's not likely to vary much. The biggest issue is that the capacitor tolerance and variation with temperature is likely to be large.

 

[Diver300]

If you are using a tranformerless supply with a capacitive current limiter and a resistive voltage divider, I don't think that you can have a full bridge rectifier. You should simply have N = 0 V. The resistive voltage divider would just have one diode to only let through the +ve half wave. The capacitive limiter would have a diode to allow the +ve half wave through, and a second diode to let the -ve half wave go to ground. The capacitor values will need to be doubled for the same current taken from the 12 V supply.

 

[AnalogKid]

I believe moving C1 to other side will not work as to to use capacitor reactance we need AC, and other side of bridge is DC

No, it isn't. It is rectified AC, an entirely different thing. The frequency spectrum is different, but the majority of the energy still is in the fundamental.

Even with the zener and capacitor in series, they present a small offset (only 7%) to the wave shape of the input voltage. The voltage across the zener comes up to 12 V in the first 6 degrees of a quarter-cycle of the sine wave, rectified or not. So for 168 degrees out of 180, the voltage waveform across the R-C network essentially is sinusoidal.

ak