3 phase 4 wire(480V) AC to 12VDC (100mA) power supply for energy metering applic

[Sultanpepper123]

I am currently working on a 3 phase energy meter project. I don't have a lot of experience in designing power supply circuits (AC/DC) let alone a 3 phase 4 wire one (L1,L2,L3 adn N) with a transformer to be galvanically isolated. Since energy measurement is being done for all three phases, the metrology (Microcontroller + metering IC) part must be able to measure energy even if two phases go out, so AC supply must come from all three phases. I've scoured the internet and found a few approaches, some used a transformer, some didn't.

Can i use something like this ( IRM-20-12 ) for the time being or if there is a better model please let me know. If the answer is yes then i think the next question is how do i connect it to all 3 phases since the power module only take Line and Neutral ?

If the answer for the previous question is no, what would be my options for a prototyping stage and another for a cost efficient and not so brutal to design ?

[dmendesf]

China is your friend:

https://m.pt.aliexpress.com/item/32824299597.html?trace=wwwdetail2mobilesitedetail&spm=a2g0s.9042311.0.0.2742b90aahVt7s

If you want to do it yourself use a Viper style power supply (but chose a IC with a higher breakdown voltage... There are >1000V options) and do 3 single phase rectifiers at the input.. so if one phase goes down you only get a slightly lower voltage and higher ripple. Be cautious, at this voltage any mistake kills.

[RIKRIK]

Not sure if this would be useful, but  if you google arduino electricity meter, there's some interesting ideas https://www.instructables.com/id/Simple-3-Phase-Arduino-Energy-Meter/

P.s I know this won't power it like you want but, three different power supplies sounds like a challenge, would running it off of mains with a back up battery be easier .You could also get a sd card module to back up all the data on.
Kind regard , rick

[NiHaoMike]

Use 3 half wave rectifiers, each must be individually fused with proper HRC fuses, but it's OK to put the EMI filtering and inrush protection after the rectifiers. You can consider using microwave oven diodes for the rectifiers, since they're fairly cheap but quite robust.

[capt bullshot]

Quite a time ago, I've seen a circuit like this in a three-phase meter (L1,L2,L3,N connected).
It uses a simple SMPS to step down and isolate the 12V suppy from an intermediate (ballpark 150V) DC voltate, derived from the line voltages by capacitive droppers having rather high series resistors. For the rather low power requirement that you have, this should also work.

Stress voltage across the diodes is reduced to said intermediate voltage, higher frequency components are dealt by the rather high impedance resistors in series with the capacitors. I don't remember the actual component values. Capacitors have to be rated for the line voltage, resistors are requires to be robust  (pulse load, max. voltage rating) and silent failing to open circuit. No fuses required, the resistors are expected to cost way less than appropriate fuses.

[Sultanpepper123]

Quite a time ago, I've seen a circuit like this in a three-phase meter (L1,L2,L3,N connected).
It uses a simple SMPS to step down and isolate the 12V suppy from an intermediate (ballpark 150V) DC voltate, derived from the line voltages by capacitive droppers having rather high series resistors. For the rather low power requirement that you have, this should also work.

Stress voltage across the diodes is reduced to said intermediate voltage, higher frequency components are dealt by the rather high impedance resistors in series with the capacitors. I don't remember the actual component values. Capacitors have to be rated for the line voltage, resistors are requires to be robust  (pulse load, max. voltage rating) and silent failing to open circuit. No fuses required, the resistors are expected to cost way less than appropriate fuses.

My apologies to you sir for not making the post clear enough , but i've just updated the picture as it is just a block diagram of the module ( it's internal components ) . Can your suggested circuit work with the inputs ?

Use 3 half wave rectifiers, each must be individually fused with proper HRC fuses, but it's OK to put the EMI filtering and inrush protection after the rectifiers. You can consider using microwave oven diodes for the rectifiers, since they're fairly cheap but quite robust.

My apologies to you sir for not making the post clear enough , but i've just updated the picture as it is just a block diagram of the module ( it's internal components ) . Being bridge rectifiers, will it work with the inputs (L and N ) ?

[ogden]

Check TI 3phase meter reference design EVM430-F6779 schematics, theres supply you are looking for.

[capt bullshot]

My apologies to you sir for not making the post clear enough , but i've just updated the picture as it is just a block diagram of the module ( it's internal components ) . Can your suggested circuit work with the inputs ?

I'm sorry, I've just thrown an idea into the discussion. Didn't read your first post carefully enough, especially the "not so brutal to design" part. Designing a circuit to work safely on a 3-phase 480V line is brutal.

[Sultanpepper123]

Check TI 3phase meter reference design EVM430-F6779 schematics, theres supply you are looking for.

Thanks checked it, as far as i can tell it's non isolated. I'd be thrilled if it had a transformer.

BTW Is there a way to connect something like the IRM-20-12 mentioned above with the 3 phase 4-wire input?

[Sultanpepper123]

My apologies to you sir for not making the post clear enough , but i've just updated the picture as it is just a block diagram of the module ( it's internal components ) . Can your suggested circuit work with the inputs ?

I'm sorry, I've just thrown an idea into the discussion. Didn't read your first post carefully enough, especially the "not so brutal to design" part. Designing a circuit to work safely on a 3-phase 480V line is brutal.

Don't be , I am aware that designing good power supplies is not easy. That's why i was looking for a ready made module like mentioned above " IRM-20-12 ", but the million dollar question is can i connect these types of modules to a three phase AC input ?.

[capt bullshot]

the million dollar question is can i connect these types of modules to a three phase AC input ?.

The correct answer is "Yes". Where can I go and collect my million dollar prize?

[capt bullshot]

Some hint: If you can't figure out from all the previous posts, just use three of these modules and tie their outputs together, decoupled by diodes. That's a way more safe way.

[Sultanpepper123]

Some hint: If you can't figure out from all the previous posts, just use three of these modules and tie their outputs together, decoupled by diodes. That's a way more safe way.

I got the idea to use the module per phase&N solution , but how can i use only one of those with some kind of clever circuitry ?

[capt bullshot]

I got the idea to use the module per phase&N solution , but how can i use only one of those with some kind of clever circuitry ?

Look up reply #3 and reply #4

[ogden]

Check TI 3phase meter reference design EVM430-F6779 schematics, theres supply you are looking for.

Thanks checked it, as far as i can tell it's non isolated. I'd be thrilled if it had a transformer.

BTW Is there a way to connect something like the IRM-20-12 mentioned above with the 3 phase 4-wire input?

Electricity meter do not need isolated supply.

Sent from my F5321 using Tapatalk

[capt bullshot]

I got the idea to use the module per phase&N solution , but how can i use only one of those with some kind of clever circuitry ?

Look up reply #3 and reply #4

Sorry for the (maybe) rude answers. I'm not going into details, since IMHO you _must_ be able to do the math to make these circuits work, and _must_ be aware of the safety related stuff. Giving you a ready made solution would be dangerous if you don't understand all the implications. So I'd recommend you to stay with the three module solution.

[Sultanpepper123]

I got the idea to use the module per phase&N solution , but how can i use only one of those with some kind of clever circuitry ?

Look up reply #3 and reply #4

Sorry for the (maybe) rude answers. I'm not going into details, since IMHO you _must_ be able to do the math to make these circuits work, and _must_ be aware of the safety related stuff. Giving you a ready made solution would be dangerous if you don't understand all the implications. So I'd recommend you to stay with the three module solution.

No not at all. I have no problem in figuring out the math behind the circuits. I'd like to sincerely thank you sir and all the fellow members here in the forums for helping me out.

[coppice]

Check TI 3phase meter reference design EVM430-F6779 schematics, theres supply you are looking for.

Thanks checked it, as far as i can tell it's non isolated. I'd be thrilled if it had a transformer.

BTW Is there a way to connect something like the IRM-20-12 mentioned above with the 3 phase 4-wire input?

Electricity meter do not need isolated supply.

Sent from my F5321 using Tapatalk

That is true in some cases, and not in others. If the meter stands alone, or is wirelessly connected, isolation is pointless, and the vast majority of products do not have isolation. However, many meters have some form of wired interface to monitoring or reading systems. These need full galvanic isolation for at least a small part of the hardware around the interface.

[ogden]

Your 3phase meter mesures voltage and current on each phase how exactly? I doubt it is done using isolated ADC frontends, that would be expensive or even dumb overengineering. As it shall be galvanically connected to neutral then theres no point to use galvanically isolated supply for it.

Sent from my F5321 using Tapatalk

[coppice]

Your 3phase meter mesures voltage and current on each phase how exactly? I doubt it is done using isolated ADC frontends, that would be expensive or even dumb overengineering. As it shall be galvanically connected to neutral then theres no point to use galvanically isolated supply for it.

Sent from my F5321 using Tapatalk

ADI makes isolated ADC front ends specifically for energy meter applications. Its no longer an expensive way to implement metering.

[ogden]

Those chips are for current shunt sensors. You are welcome to show meter reference design where voltage channel is isolated. Perhaps I missed it

Sent from my F5321 using Tapatalk

[coppice]

Those chips are for current shunt sensors. You are welcome to show meter reference design where voltage channel is isolated. Perhaps I missed it

Sent from my F5321 using Tapatalk

The ADI chips provide two isolated sigma delta modulator channels. One for a voltage divider sensor, and a high gain one for a current shunt.

The relevant device numbers are ADE7978, ADE7933 and  ADE7932.

[ogden]

Yes you are right. ADE79xx are nice isolation ADC's. Thank you.

Sent from my F5321 using Tapatalk

[Sultanpepper123]

Quite a time ago, I've seen a circuit like this in a three-phase meter (L1,L2,L3,N connected).
It uses a simple SMPS to step down and isolate the 12V suppy from an intermediate (ballpark 150V) DC voltate, derived from the line voltages by capacitive droppers having rather high series resistors. For the rather low power requirement that you have, this should also work.

Stress voltage across the diodes is reduced to said intermediate voltage, higher frequency components are dealt by the rather high impedance resistors in series with the capacitors. I don't remember the actual component values. Capacitors have to be rated for the line voltage, resistors are requires to be robust  (pulse load, max. voltage rating) and silent failing to open circuit. No fuses required, the resistors are expected to cost way less than appropriate fuses.

Just a followup for people who may end up in my situation, i've read and quite alot of implementations using this circuit. I'd like to point out that maybe the diode between line and neutral are reversed maybe ? .

Regards


Reference : http://www.ti.com/lit/df/tidr430/tidr430.pdf
My apologies to you sir for not making the post clear enough , but i've just updated the picture as it is just a block diagram of the module ( it's internal components ) . Can your suggested circuit work with the inputs ?

Use 3 half wave rectifiers, each must be individually fused with proper HRC fuses, but it's OK to put the EMI filtering and inrush protection after the rectifiers. You can consider using microwave oven diodes for the rectifiers, since they're fairly cheap but quite robust.

My apologies to you sir for not making the post clear enough , but i've just updated the picture as it is just a block diagram of the module ( it's internal components ) . Being bridge rectifiers, will it work with the inputs (L and N ) ?

[splin]

An important aspect that you haven't stated is where this meter is to be installed as it very significantly affects the surge/transient voltage ratings the PSU has to withstand. https://www.tequipment.net/cat-ratings-explained/] [url]https://www.tequipment.net/cat-ratings-explained/[/URL] is a good guide.

Being three phase 480V it is likely to be cat II, cat III or cat IV 300V, as defined in IEC 61010-1. Cat II means it has to withstand 2.5kV impulses from a 12 ohm source. Cat III and IV withstand voltages are 4 or 6kV impulses from a 2 ohm source respectively which significantly limits the choice of off the shelf PSUs and they will be rather more costly than the common CAT II models. This assumes you need to be compliant with 61010-1 or equivalent of course but it would be risky, legally and ethically, to ignore the standard.

If you don’t need an isolated supply things are easier; ie. if your meter does not have any accessible ports including USB, serial or parallel I/O etc. Either way it still has to be safe when subject to the test voltages without catching fire or resulting in an uncontained explosion of one or more components – there is a lot of energy available from a 6kV 2ohm source.

You could use a lower category PSU and add additional protection components (the IRM-20-12 is spec’d as 3kV AC isolation). It would be interesting to know how cheaply a CAT III and IV 300V low power supply can be implemented but I have to agree that avoiding the need for HRC fuses (very expensive) has to be a good starting point, but 6kV+ capacitors and diodes aren’t cheap either. Diodes have a tendency to fail short circuit which won’t help in creating a low cost, safe design. I’m no expert here so hopefully someone experienced in the field will help out – preferably without the patronising ‘leave it to the professionals’ type responses beloved of some sparkies/gas engineers.

Note that those test voltages apply where the insulation is defined as 'BASIC INSULATION and SUPPLEMENTARY INSULATION' (ie. double insulation). If it is 'REINFORCED INSULATION' (eg. a multimeter test lead), those test voltages increase to 6.4kV and 9.6kV and creepage and clearance spacings double. I doubt that this will be an issue as I assume most equipment uses double insulation.