I think you quoted the wrong text. My response to the OP was not about capacitors.

Indeed I did goof! I was intending response to the TS. The fact is that operating above the rated maximum voltage is asking for trouble. And changing over to a "Y" arrangement half wave supply from three phases is a simple revision. Leaving parts out is simple, adding the neutral may be a bigger deal .And stepping down from 400 volts to 9 volts just does not seem like the best plan.
I think there was a thread about doing that with resistors a short while back.

 

Another observation of the circuit, capacitor C26 (2n2F, 2kV) is bridging the reinforced isolation barrier between primary and secondary circuits – this must be a Y1 rated safety capacitor meeting IEC 60384-14; the voltage rating of 2kV would suggest otherwise.

Actually, it's not. It is connected between V+ and earth. It's not a Class-II device. Class X will suffice.

 

But that's not a long term solution. Just can't label "THIS DEVICE MUST BE PROTECTED WITH AN SPD FOR RELIABLE OPERATION" on the package!

These products are often installed at very remote locations in India such as at the centre of a large paddy field or so where lightning induced voltages are really harsh where they seldom get many paths to earth to settle down.

It would seem a logical installation requirement for those types of locations. You, the installer should provide a safe path for lightning induced energy to flow instead of trying to stop it totally (yes, there still needs to be robust measures also in the electronics) inside the electronics.

Sure, you can label that as a recommendation. That's what is done on the majority of remote or off-grid solar panel installations.
https://info.littelfuse.com/surge-protection-devices-for-solar-applications

 

Actually, it's not. It is connected between V+ and earth. It's not a Class-II device. Class X will suffice.

Confusingly the earth/ground symbol is used in the circuit diagram to represent the secondary 0V (and not earth/ground), which is separated from real earth/ground (chassis/PE) by an inductor (L3) – therefore my statement that the capacitor C26 must be a Y1 class safety capacitor holds true.

 

NO!! I was not mentioning those capacitors filing .I m describing the voltage cross the input to the switching regulator, which can be quite bit above the 400 volts stated. That voltage can easily exceed 450 volts if the input is 400 volts between phases.
What I suggested was changing the supply to lower voltage arrangement that used the neutral connection.
That is clearly stated in the second paragraph of post #13.
If byou do not understand a half wave rectifier perhaps someone else will explain it. That arrangement will require neutral connection.

Cannot rearrange the supply to lower voltage arrangements that includes neutral as the utility company doesn't supply power to many of these locations with a neutral to prevent people misusing these government subsidised connections for other purposes (usage of subsidised connection for personal uses / theft). Neutral connection is terminated at the sealed energy meter and is not made available from there. Any connections we tap from the input (including neutral) of energy meter is illegal.

Also the voltage to input switching regulator is linearly brought down to 318V by the IRF830, P6KE300A and P6KE18A TVS diodes arrangement. So the input switching regulator (Viper22A) is never presented with a voltage beyond rating. To substantiate further, these devices can also work in single phase where to out of the three phase inputs we can connect one input to any phase and any other input to neutral leaving one input unused essentially making it a 230Vac powered power supply. We are facing the same component damages at such sites as well. So this cannot be an over voltage issue.

 

Confusingly the earth/ground symbol is used in the circuit diagram to represent the secondary 0V (and not earth/ground), which is separated from real earth/ground (chassis/PE) by an inductor (L3) – therefore my statement that the capacitor C26 must be a Y1 class safety capacitor holds true.

Sorry that my schematics mislead you there. The ground symbol with the '0' connected to L3 is the reference earth ground potential reference to a differential opamp voltage amplifier circuit that connects to ADC to measure phase voltages. Has got nothing to do with the secondary ground. You may just pretend that L3 does not exist.

The secondary ground is not hard etched on the PCB to earth, but connected to earth using a separate wire soldered from earth to secondary ground during production. This is why chassis ground and secondary ground has different symbols. Regarding C26, it was provided in schematics to use the design in class II applications if required. Currently it is not placed in production and the secondary ground is connected to earth as described above.

 

Found a point in this article here that states

Putting in an LC filter works well for a DM source, as it simply presents a 'wall' of impedance which serves to block the DM emissions from entering the mains lines, and re-routes them accordingly. But this strategy is not by itself going to be effective for CM noise, whose source in a typical power supply behaves more like a current source. Current sources demand to keep their current flowing, and can surmount any impedance 'wall' we may throw at them (if there is no other route available). That is why in a power converter, we have to place a freewheeling diode otherwise the inductor current would punch a hole in our Fet the moment we tried to turn it OFF. We also know that pure reactances cannot dissipate any energy, only resistances do.

All this leads to the following strategy when dealing with CM emissions— a) we must allow it to flow b) we can re-route it thereafter by using impedance mismatching c) finally we do need to dissipate this energy by using resistive elements.

To me that makes sense. Similar to how a current transformer works, that is it raises it's voltage across it terminals so high as to make current through its sec winding become equal to primary current divided by number of turns - either through the burden resistance or arcing through its terminals - The reason why they should be handled with utmost care.

During a lightning event, the primary current (current that flows from cloud to cloud or cloud to earth or whatever) is so high that the public electric lines acts as a current transformer secondary, pushing its output voltage high enough to break any impedances on its path to complete it's flow.

So I think varistors at the input connected to protective earth from each phase followed by common mode filter are the only option we got.

Also as @Hymie suggested, we will also put in an RCD clamp to increase switching rise time.

 

A further observation on the switch mode PSU design is that normally 3 phase is used where there is insufficient power available from a single phase (without up-rating the supply connection), but here it is being used for redundancy, in that should any one of the three phases go down, the supply will still operate normally.

However it should be possible to redesign the front end to include a neutral connection such that should two of the three phases go down or the neutral connection lost, then normal operation continues – this would make the product even more resilient to supply losses, particularly in locations where such power losses are not uncommon.

 

A further observation on the switch mode PSU design is that normally 3 phase is used where there is insufficient power available from a single phase (without up-rating the supply connection), but here it is being used for redundancy, in that should any one of the three phases go down, the supply will still operate normally.

However it should be possible to redesign the front end to include a neutral connection such that should two of the three phases go down or the neutral connection lost, then normal operation continues – this would make the product even more resilient to supply losses, particularly in locations where such power losses are not uncommon.

Thanks for the suggestion. But these are three phase motor control systems. So since the motor will not work without atleast 2 phases available (which should be avoided), it's safe for the device to power off when only one phase is available. Plus there is battery backup for the device built-in as well to remain powered during power loss for upto a day. Also unavailability of neutral at some locations as explained in my previous replies is also a reason why we decided to keep the system neutral free

 

Found a point in this article here that states



To me that makes sense. Similar to how a current transformer works, that is it raises it's voltage across it terminals so high as to make current through its sec winding become equal to primary current divided by number of turns - either through the burden resistance or arcing through its terminals - The reason why they should be handled with utmost care.

During a lightning event, the primary current (current that flows from cloud to cloud or cloud to earth or whatever) is so high that the public electric lines acts as a current transformer secondary, pushing its output voltage high enough to break any impedances on its path to complete it's flow.

So I think varistors at the input connected to protective earth from each phase followed by common mode filter are the only option we got.

Also as @Hymie suggested, we will also put in an RCD clamp to increase switching rise time.

OK, in that case, to prevent the fast transients from entering, the portion of the protection arrangement should be before the rectifier portion. That means immediately after the three 100 ohm resistors, R76, R60, and R48. In addition, considering that the power provided by the 12 or 9 volt supply is quite low, although we do not have the exact number, those resistors could be a higher value.
The series inductance that I am suggesting as an input filter to block the fast transients from lightning would need to present an impedance of at least 1000 ohms at a frequency of 1000 Hz. The filter will need to be before the three capacitors connected between the three phases.
It is not reasonable to presume that a shunt impedance will do so much to reduce the fast transients, so blocking them with a series impedance is a more reasonable approach.
And if the power utility does not provide a neutral conductor, can it be assumed that the neutral is actually tied to a true earth "ground"???

 

OK, in that case, to prevent the fast transients from entering, the portion of the protection arrangement should be before the rectifier portion. That means immediately after the three 100 ohm resistors, R76, R60, and R48. In addition, considering that the power provided by the 12 or 9 volt supply is quite low, although we do not have the exact number, those resistors could be a higher value.
The series inductance that I am suggesting as an input filter to block the fast transients from lightning would need to present an impedance of at least 1000 ohms at a frequency of 1000 Hz. The filter will need to be before the three capacitors connected between the three phases.
It is not reasonable to presume that a shunt impedance will do so much to reduce the fast transients, so blocking them with a series impedance is a more reasonable approach.
And if the power utility does not provide a neutral conductor, can it be assumed that the neutral is actually tied to a true earth "ground"???

So are you suggesting a 159mH ((1000 ohms)/(2*pi*1000Hz) three winding common mode choke immediately after R76, R60 and R48? That would be a pretty bulky one I guess!

Regarding your mention about the neutral, I didn't understand the question quite well. But it's not that the utility company does not provide neutral, but that they don't give it for use in the pump room. But yes it's tied to earth ground over an earth pit at a nearby electric post

 

These products are often installed at very remote locations in India such as at the centre of a large paddy field

But yes it's tied to earth ground over an earth pit at a nearby electric post

So is your location India?
In N.A. and most jurisdictions, the Neut. can only be referenced to earth ground at the power supply source or entry to the installation.

 

So are you suggesting a 159mH ((1000 ohms)/(2*pi*1000Hz) three winding common mode choke immediately after R76, R60 and R48? That would be a pretty bulky one I guess!

Regarding your mention about the neutral, I didn't understand the question quite well. But it's not that the utility company does not provide neutral, but that they don't give it for use in the pump room. But yes it's tied to earth ground over an earth pit at a nearby electric post

It would not need to be a 3-circuit common mode choke., probably individual chokes will be more likely available. And the exact inductance value is not critical, because the purpose is simply to present a higher impedance to the fast transients. The current rating for the device needs to be adequate for the application. What is the anticipated current draw at the three-phase connection??

 

I normally use something like this (sized to the correct current draw of the device) for a 3-phase utility noise filter.

It works for what it's designed for but it's NOT a 3-phase lightning surge suppressor (a current source that will transform that energy to whatever voltage is needed to complete the circuit if you try to just block it) that in remote installations, should be physically separated and grounded where it's 'safe' to shunt energy.

https://www.eaton.com/us/en-us/prod...e-it-power-distribution/surge-protection.html

 

Something else you may or may not have considered in relation to the battery within the unit – charge control is provided by transistors Q5 & Q7 and associated components, however with faults within this circuitry (say diode D25 short circuit) the full V SMPS will be permanently applied across the battery. With capacitor C19 short circuit the battery will be subject to a short circuit; you should ensure that neither of these fault conditions causes a hazard as a result of abusing the battery.

 

What is this low voltage power source powering? And how much power is actually required? For small power demands a small transformer makes sense. And if the small transformer had adequate insulation between primary and secondary it could resist any damage from the fast transients.
How isolated is whatever load is being powered by this system??

 

Something else you may or may not have considered in relation to the battery within the unit – charge control is provided by transistors Q5 & Q7 and associated components, however with faults within this circuitry (say diode D25 short circuit) the full V SMPS will be permanently applied across the battery. With capacitor C19 short circuit the battery will be subject to a short circuit; you should ensure that neither of these fault conditions causes a hazard as a result of abusing the battery.

The battery pack is a 7.4V li-ion pack with inbuilt charge protection circuitry. So over voltage, charge current limiting and over charging are taken care of by it. The onboard current limiting by Q5 and Q7 are provided just not to overload the smps with too high a charging current by the battery pack

 

So is your location India?
In N.A. and most jurisdictions, the Neut. can only be referenced to earth ground at the power supply source or entry to the installation.

You could be right. But wouldn't it be a good thing to do connecting neutral to earth at the electric post?

What is this low voltage power source powering? And how much power is actually required? For small power demands a small transformer makes sense. And if the small transformer had adequate insulation between primary and secondary it could resist any damage from the fast transients.
How isolated is whatever load is being powered by this system??

We need the system to be powered whenever atleast two out of the three phases are available, even if the battery is damaged / fails. So transformer is not an option. Also, the peak power requirement is 20W

 

OK, the "system" has a peak power demand of 20 watts. How long will that peak demand be? Hours or seconds?
And two transformers connected across two of the three lines will deliver power if any two lines are active, O there is one possible scheme to have isolated power from any two lines. And if there are other three phase loads on the same side of the failure you could possibly get useful power if only one phase were active.
And the transformers could be well enough insulated to avoid any damage to the circuits from fast transients. And it may be possible to create a single transformer for the application.

 

What does the equipment do?