Discuss this simple way to realise my 220VAC/lowDC converter? Wrong or right?

[6716914]

I am a chip designer. I have some questions about power converter.

Now I want to use a simple device to convert the 220Vac to a low DC (1~3V current is around 10mA). The efficiency demand is not very high, more than 10% is OK. I want to design this power supply to a chip.

First idea is to use a linear capacitor to get the 2V AC, two capacitors, 120nF and 12uF, and then, I use a 1Kohm parallel to 12uF, then the load's output voltage is around 2V ac. Since the caps consume little energy, I think the efficiency will more than 10%. But I dont see peope mention this, and never see someone use this. I am not an expert for this. But according to my simulation result, it works. Is this methold reasonable? If it is, then I will put the 120nF and 12uF caps outside the chip, and use my nm Si process chip to convert this 2AC to the DC voltage I need.

Second idea, is about the transformer.My idea is to use a transformer to convert a 220V ac to a fairly low voltage (2V~12v), and then I will use a nm process chip to design the low voltage ac to the low voltage dc I really need. So I am searching alot. I am very interested in the microtransformer design or the PCB stack transformer design. I know the main frequency is very low 50Hz, so I am not sure to design it in that way.

BTW, Switch mode, I already think about it, because it needs lots big capacitors, and I dont think it is what I want.

This converter should be plugged into the wall, and convert to a low DC voltage to drive my CMOS device.

 

[kripacharya]

I am a chip designer. I have some questions about power converter.

Now I want to use a simple device to convert the 220Vac to a low DC (1~3V current is around 10mA). The efficiency demand is not very high, more than 10% is OK. I want to design this power supply to a chip.

First idea is to use a linear capacitor to get the 2V AC, two capacitors, 120nF and 12uF, and then, I use a 1Kohm parallel to 12uF, then the load's output voltage is around 2V ac. Since the caps consume little energy, I think the efficiency will more than 10%. But I dont see peope mention this, and never see someone use this. I am not an expert for this. But according to my simulation result, it works. Is this methold reasonable? If it is, then I will put the 120nF and 12uF caps outside the chip, and use my nm Si process chip to convert this 2AC to the DC voltage I need.

Second idea, is about the transformer.My idea is to use a transformer to convert a 220V ac to a fairly low voltage (2V~12v), and then I will use a nm process chip to design the low voltage ac to the low voltage dc I really need. So I am searching alot. I am very interested in the microtransformer design or the PCB stack transformer design. I know the main frequency is very low 50Hz, so I am not sure to design it in that way.

BTW, Switch mode, I already think about it, because it needs lots big capacitors, and I dont think it is what I want.

This converter should be plugged into the wall, and convert to a low DC voltage to drive my CMOS device.

Both ideas will work, assuming you can get a transformer for that ratio. In fact first method is used for simplistic transformerless power supplies (look it up) for low currents. However note that isolation from your mains supply is not there, and hence it's a dangerous/ hazardous method of doing it. Also note that 220Vac is the rms value, and the peaks will be 220x 1.414.

 

[nikhilt007]

hai
here you mentiond that you need a small current about 10ma so you should use transformerless powersupply
better than using transformer since transformer is mostly designed to get high current in 1a ,2a etc other wise you have to look for transformer ratio
since you need very low current it is better to use resistive transformerless power supply
try for tthese
and be careful you are directly converting 230ac with out ic

i suggest to use resistive power supply