Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Is this Cap-Drop power supply truly reversed ?!

Status
Not open for further replies.
hm_fa_da

hm_fa_da

Full Member level 5
Full Member level 5
Joined
Sep 16, 2003
Messages
287
Helped
10
Reputation
20
Reaction score
4
Trophy points
1,298
Visit site
Activity points
3,217
Hello,

I found a Cap drop power supply Schematic in which 220V AC LINE is connected directly to 3.3V VCC of output voltage ( 220 to 3.3V ):
positive.png

for a reason i need to reverse out put so GND be connected to 220V AC LINE !
so i changed schematic to this :
negative.png

is the second schematic truly changed ?!!

Thanks & Regards.
 

Your schematic appears more complicated than it needs to be. It may be possible to simplify, so that the TL431 is the only voltage regulation you need.

Can you verify that the dropping capacitor, 100 pF, can pass enough current to drive a circuit containing capacitors 100 uF, and a 1000 uF?
 

sorry, i missed changing the value in schematic, the true value for the input cap is 200nF/275V .

the first schematic is obtained from an application note, the second one is what i changed to reverse output voltage ...
 

I also don't get the purpose of the multiple transistors, diodes and capacitors. Can you link the application note?

Presumed the original circuit works, flipping all component's polarity should do. If the input TVS is actually unipolar as the symbol says, it must be flipped, too.

The DC output is floating and must be isolated from any external circuit. In so far you don't necessarily need to change the circuit, just connect the load differently.
 

Thanks for the informative link. I didn't notice before that one line input is directly connected to the output side, as required for a meter circuit. In so far your modification makes sense when you need a different output polarity.
 

i have another question too, apart from polarity, i found this cap drop power supply design from ST too :
ST-SUPPLY.png
which uses both positive and negative wave of input 220V, like the maxim design but ST uses less components.
i saw ST design in a meter application note, not talking anymore about power supply output capacity ( like in maxim application note ).
now do you think this power supply be suitable for me too ?
my circuit draws 15ma maximum at 3.3V.
 

There is one important difference between your schematic and the one in the data sheet - you show a ground connection, the data sheet doesn't. Also note that the supply is intended to be used to power line metering circuits which are by design completely isolated from any outside circuitry. They typically reside inside the box of the electricity meter which in turn is secure and isolated from outside contact. It is NOT intended for general use and you can NOT connect a ground to it. The circuit does not, and can not isolate you from the line voltage so please treat it with extreme care and never use it to power anything you can touch.

Brian.
 

betwixt said:
There is one important difference between your schematic and the one in the data sheet - you show a ground connection, the data sheet doesn't. Also note that the supply is intended to be used to power line metering circuits which are by design completely isolated from any outside circuitry. They typically reside inside the box of the electricity meter which in turn is secure and isolated from outside contact. It is NOT intended for general use and you can NOT connect a ground to it. The circuit does not, and can not isolate you from the line voltage so please treat it with extreme care and never use it to power anything you can touch.

Brian.
Click to expand...

the ground symbol in my schematic doesn't mean Earth connection, and it is completely isolated from outside too ( in my usage ).
i just have 2 questions, in first schematic, did i reverse the polarity true ?
in the last post before this, cap drop power supply is designed by ST, what's practical differences between this and first schematic ?!
 

I think the intention of the first design is to improve the symetry of current it draws but both designs should function equally. You can even take the AC from after the capacitor straight to a bridge then the regulator if cost has to be absolute minimum. Both the designs have some advantage in that their low side output voltages are almost at neutral potential while a bridge would float at a higher voltage.

I think in the ST design, I would be inclined to add a high value resistor (~1M) across C18 just in case the input became disconnected. There is a slight risk that the AC input connections could retain voltage if C18 remained charged up. Also note that R28 should be a safety type (fusible) in case C18 fails.

Brian.
 

The ST design is untested as shown because Vbe negative AC swing would be excessive(>|-300Vp|) . THis can easily be corrected with a 500V series silicon diode but the 15k base resistor will dissipate less than 5 watts, so must be selected appropriately. All caps xxx nF caps are plastic hi V types. THe programmable Zener is set to 3.7V for float charging a LiPo battery or an LDO as shown and it pumps up to +3.7 in less than 1 second.

Otherwise it is a good design. I'm just surprised that ST released it, as it is.

- - - Updated - - -

THe 15k resistor can be increased to 100k,1W and reduced power if a 1nF film cap is put in parallel., However the Vce must also be rated for 400V so the transistor must be changed .

Design is not looking so great now. with 3mA out at 3.7 from Zener with 30mA in at 230Vac 50Hz. Watt's that? 10mW out for about 5 VAR input?

- - - Updated - - -

Unless you have a lot of experience, I would avoid of offline regulator DIY designs. There are many pitfalls with line transients and overvoltage.

For $2 it isn't worth the challenge unless you have some other motive.
https://www.monolithicpower.com/DesktopModules/DocumentManage/API/Document/getDocument?id=3454

https://www.monolithicpower.com/DesktopModules/DocumentManage/API/Document/getDocument?id=3562

$2 plus a few passives

Universal AC Input (85VAC-to-305VAC)
• Inductor-Less
• Less than 100mW Standby Power
• Excellent EMI
• Low BOM Cost
• Smart Control to Maximum Efficiency
• Adjustable Output Voltage from 1.5V to 15V, 20~30mA
• Good Line and Load Regulation
• Thermal Shutdown Protection
• Short-Circuit Protection
• Provide Power-Good Signal
APPLICATIONS
• Wall Switches and Dimmers
• Z-Wave Device and ZigBee Device for Home Automation
• Standby Power for General Off-line Applications

Unless you have a burning urge to burn parts or your load or your finger, leave offline regulator designs to the experts.

- - - Updated - - -

TO answer your original question;

THe Maxim design is not reversed but rather the output common ground is 3.3V below the Line voltage . THis might? be ok for low frequency Teridian meters but terrible for microwave radios built into meter as the local ground voltage modulates the stray capacitance at microwave frequencies and implies an antenna load current if there is a radio link with any proximity to the earth grounded meter structure.

Not only is your version wrong but your input cap is in pF instead of nF, and it is NEUTRAL not NULL.

I have 5 yrs experience with Power Meters , automated meter reading with ISM 1GHz radios.

The worst meters were only one ancient style of Westinghouse meters that used a plastic in the magnetic air bearing called Meltric? that was lossy and high K and whenever the disk rotated , energy at 1GHz was reflected and amplitude modulated by the resonant structure of the disk coupled by a tiny piece of plastic, about 50pF in the bottom of the magnetic air bearing spindle. It took me a year to figure that out in my spare time as Op Mgr, while the RF Engineers could not including my boss who had a PhD and much more experience than I. (actual R&D time 8 hrs)
 
Last edited:
Thanks for your helps,

Not only is your version wrong but your input cap is in pF instead of nF, and it is NEUTRAL not NULL.
Click to expand...

i just put the value wrong in schematic, that cap is 220nF same as maxim design.
my purpose is to design almost same meter as maxim did but with other company (Atmel) chips which consume same low power and there is no radio link too.

and about MPL100, the price is good for mass production ( about 0.6USD in digikey ), but there is a problem in the schematic recommended in MPL datasheet, that is they used Diode bridge in input of the AC voltage !
all the reason ST and maxim designed that circuits with transistor for negative cycle is because of what i said.
the bridge can't be used in input side !
the meter ICs use Shunt to measure load current so there is not isolation to Line voltage, line is connected directly to 1 input of differential ADC inputs on meter IC. so the line should be used as GND of whole circuit to avoid high voltage on meter pin vs GND.

here is the shunt connection schematic :
shunt.png

That's why i was trying to revers the Maxim design because they connected Line to Vcc3.3V but i needed Line to be connected to GND. ( the voltage on differential input pins of the ADC chip is -1V to VCC ) so the the GND net should be at 0V.

now a question, can MPL be used with half cycle rectifier Diode ?! ( 1 diode only )
what changes should be applied to do so ?
 

SunnySkyguy said:
The ST design is untested as shown because Vbe negative AC swing would be excessive(>|-300Vp|) . THis can easily be corrected with a 500V series silicon diode but the 15k base resistor will dissipate less than 5 watts, so must be selected appropriately. All caps xxx nF caps are plastic hi V types. THe programmable Zener is set to 3.7V for float charging a LiPo battery or an LDO as shown and it pumps up to +3.7 in less than 1 second.

Otherwise it is a good design. I'm just surprised that ST released it, as it is.

- - - Updated - - -

THe 15k resistor can be increased to 100k,1W and reduced power if a 1nF film cap is put in parallel., However the Vce must also be rated for 400V so the transistor must be changed .

Design is not looking so great now. with 3mA out at 3.7 from Zener with 30mA in at 230Vac 50Hz. Watt's that? 10mW out for about 5 VAR input?

- - - Updated - - -

Unless you have a lot of experience, I would avoid of offline regulator DIY designs. There are many pitfalls with line transients and overvoltage.

For $2 it isn't worth the challenge unless you have some other motive.
https://www.monolithicpower.com/DesktopModules/DocumentManage/API/Document/getDocument?id=3454

https://www.monolithicpower.com/DesktopModules/DocumentManage/API/Document/getDocument?id=3562

$2 plus a few passives

Universal AC Input (85VAC-to-305VAC)
• Inductor-Less
• Less than 100mW Standby Power
• Excellent EMI
• Low BOM Cost
• Smart Control to Maximum Efficiency
• Adjustable Output Voltage from 1.5V to 15V, 20~30mA
• Good Line and Load Regulation
• Thermal Shutdown Protection
• Short-Circuit Protection
• Provide Power-Good Signal
APPLICATIONS
• Wall Switches and Dimmers
• Z-Wave Device and ZigBee Device for Home Automation
• Standby Power for General Off-line Applications

Unless you have a burning urge to burn parts or your load or your finger, leave offline regulator designs to the experts.

- - - Updated - - -

TO answer your original question;

THe Maxim design is not reversed but rather the output common ground is 3.3V below the Line voltage . THis might? be ok for low frequency Teridian meters but terrible for microwave radios built into meter as the local ground voltage modulates the stray capacitance at microwave frequencies and implies an antenna load current if there is a radio link with any proximity to the earth grounded meter structure.

Not only is your version wrong but your input cap is in pF instead of nF, and it is NEUTRAL not NULL.

I have 5 yrs experience with Power Meters , automated meter reading with ISM 1GHz radios.

The worst meters were only one ancient style of Westinghouse meters that used a plastic in the magnetic air bearing called Meltric? that was lossy and high K and whenever the disk rotated , energy at 1GHz was reflected and amplitude modulated by the resonant structure of the disk coupled by a tiny piece of plastic, about 50pF in the bottom of the magnetic air bearing spindle. It took me a year to figure that out in my spare time as Op Mgr, while the RF Engineers could not including my boss who had a PhD and much more experience than I. (actual R&D time 8 hrs)
Click to expand...

Hello again,

about the ST design, they not just released it, they also made it ( as pictures of the made board is shown in document ) "STEVAL-IPE012V2: single-phase energy meter with 80 A maximum
current based on the STPM10 metering IC and STM8L152C6 MCU".

and about the maxim design, you said my version is wrong ! ( the 220PF capacitor was a mistake written value in schematic, it is 220NF ) , i just found a schematic from Atmel in document AVR1631 schematic in which they exactly used a schematic like me to revers the output voltage ( to make 3.3V higher than GND ( Line ) )
here is atmel schematic:
Atmel.png

however i also think ST design lacks a Zener or TVS diode after drop cap in input to limit voltage !

but i don't understand why you told my version is wrong ?!
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top