Common mode surge currents in Flyback SMPS?

[cupoftea]

Hi,
Page 17 of the innoswitch3-EP datasheet says that due to “common mode surge currents”, the Y capacitor should be connected from Primary High voltage bus to output positive.
That sounds like a good place to connect the Y capacitor, ….but rather, for the reason that both those places are relatively quiet nodes (not switching nodes)
If the output is not earthed, what “common mode surge currents” will flow through the Y capacitor?
Its said that these “common mode surge currents” need routing away from the primary side PWM controller…..but how often do such “common mode surge currents” flow?
The section goes on to state that……

“– if an input pi-filter (C, L, C) EMI filter is used then the inductor in the filter should be placed between the negative terminals of the input filter capacitors”.

…so they are saying this presumably because otherwise, if the inductor is placed in the high side, then it will mean more common mode emissions?...but then what about the common mode choke in the AC filter section?…..so surely it doesn’t matter whether that filter inductor goes in the high or low side?

Innoswitch3-ep datasheet

https://www.power.com/sites/default/files/documents/innoswitch3-ep_family_datasheet.pdf

 

[Easy peasy]

you need to read up on the artificial hand test ....

 

[cupoftea]

Thanks, it looks like like they are referring to general common mode emissions, rather than surges.

--- Updated Apr 16, 2023 ---

A common mode "surge" is what you get eg in a lightning strike?

 

[Easy peasy]

please read post #2 again

--- Updated Apr 16, 2023 ---

--- Updated Apr 16, 2023 ---

When testing a plastic case psu with no earth - the testing lab will wrap the psu in foil - earthed - to simulate some one holding it - this usually gives the worst case emissions ( mainly CM ) as the cap to earth is now raised.

 

[cupoftea]

When testing a plastic case psu with no earth - the testing lab will wrap the psu in foil - earthed - to simulate some one holding it - this usually gives the worst case emissions ( mainly CM ) as the cap to earth is now raised.

Thanks, we have seen our competitior has one 8W isolated offline SMPS in a plastic housing which has vent holes in it for the fan. It has no earth connection, How would they wrap this in metal foil?..it would obscure the fan's vent holes?

 

[Easy peasy]

Correct

 

[D.A.(Tony)Stewart]

the Y capacitor should be connected from Primary High voltage bus to output positive.

I believe C10 is the Y cap referred to connect to positive or negative bulk out.

It is my understanding this cross connection to negative cancels out the forward coupling transformer leakage capacitance by design to reduce leakage. But earth grounding the secondary will stress the primary insulators even more from transients.

 

[cupoftea]

Correct

Thanks, but if the vent holes are obscured the product will overheat.
Surely if you have vent holes you dont get the "artifical hand test"?
Isnt the "artifical hand test" only for medical supplies?

 

[Easy peasy]

Depends how long you run it for under emc test. - 60sec for a sweep ?

EMC tests are always worst case - hence the hand / foil test

--- Updated Apr 19, 2023 ---

Also - if it overheats and the internal protection operates - this may be a real world worst case condition for EMC - due to any weird switching action under over heat - and thus need testing.

 

[cupoftea]

Thanks, talking "worst case", as you know, its usually full power. But it amazes me that many EMC testers dont actually monitor the power to the DUT whilst doing conducted EMC testing...once had a battery charger product under EMC test...they allowed us to EMC test it with full batteries...so the charger wasnt working...and it was just on a few watts of standby power...whereas with the charger going it would have been at 150w.

Also there are loads of mains powered products in plastic cases with vent holes, and i am amazed they make them all go through the artifical hand test?.......i mean, if it needs testing whilst in overheat, then that should surely be a separate test. I have never heard of a conducted emc test needing to be carried out with the product entering into one of its fault modes.

 

[Easy peasy]

The EU standards are quite clear, stating worst case for emc, the manufacturer is bound by the statute to ensure the emc tests are performed at worst case

if a complaint is lodged and upheld - the right to market that product in the EU would be revoked

similarly for FCC in the States, and the equivalent outfit in Canada ...

Since the UK is now out of the EU - you had better check their laws

--- Updated Apr 19, 2023 ---

any domestic psu with holes could be bound up in a blanket in the real world and have to rely on its internal overtemp protection - similarly for an industrial psu left to operate in a closed box.

 

[cupoftea]

Yes, thankyou...sorry to sound cynical here .....So if you go to the official conducted EMC lab...and they do the test, and give you a certificate that says your product has passed...do you then go back to them and say "but i'm not sure if you were strict enough...can you repeat please?". I think most places just bag the cert and run...if someone complains...you have an official certificate.

 

[cupoftea]

Also, if there is this "artificial had test", then surely everyone would build their "offline SMPS in plastic" with an earth connection, and also put a metal earthed enclosure inside the plastic enclosure?......which in fact, gets done anyway, (metal enclosure inside the plastic) as no offline SMPS will ever pass radiated emissions if its in a totally plastic enclosure...)

When testing a plastic case psu with no earth - the testing lab will wrap the psu in foil - earthed - to simulate some one holding it - this usually gives the worst case emissions ( mainly CM ) as the cap to earth is now raised.

Thanks, so does that mean that all you have to do to get out of the "artificial had test", is have an earth connection to the plastic cased offline SMPS?.......then you dont even need to do anything with the earth connection because theyll just see that their is one, and not do the Artificial hand test"?

 

[cupoftea]

When testing a plastic case psu with no earth -
the testing lab will wrap the psu in foil -
earthed - to simulate some one holding it -
this usually gives the worst case emissions ( mainly CM )
as the cap to earth is now raised.

Thanks, I see that test shows EN61000-4-6, which pertains to “conducted and radiated immunity” only. I wonder if this “artificial hand” gets used for eg “standard conducted and radiated emissions”?

The whole thing seems very odd though…..i mean, if this “artificial hand test” really gets done, then surely every manufacturer would take the mains earth wire to the offline SMPS, and use it for connection of Y capacitors, so as to increase common mode filtration, and thence, assist them in passing conducted EMC when this “aritifcial hand” test is done?

Wrapping a handheld product, (which is powered by an offline SMPS) in metal foil, when that handheld product has internal cooling fans , and relies for its operation on blown air coming from vent holes in the product case….which will now be blocked by the “artificial hand” sounds bizarre.

There must be some regulatory clause which means “artificial hand” does not get used in such cases?…because the product simply could not operate in these cases….and so could not be tested.

 

[cupoftea]

This shows the "artifical hand"

..seen at 1:42...however, it barely wraps round one little bit of the hairdryer base, and does not cover the vent holes. However, the act of doing it is going to worsen the CM emissions, so its odd that all hand held devices arent supplied by SMPS's with earth taken to them, ...then the CM could be reduced by having y caps to earth....in conjunction with CM choke.

 

[D.A.(Tony)Stewart]

This shows the "artifical hand"

..seen at 1:42...however, it barely wraps round one little bit of the hairdryer base, and does not cover the vent holes. However, the act of doing it is going to worsen the CM emissions, so its odd that all hand held devices arent supplied by SMPS's with earth taken to them, ...then the CM could be reduced by having y caps to earth....in conjunction with CM choke.

The hand is like the Y caps from CM to earth (body) so the foil is diverting some conducted CM emissions to the foil capacitance to Gnd.

So your assumption that conducted emissions will worsen is false. You may be thinking of Radiated emissions.

 

[cupoftea]

Thanks, unfortunately, as you know, the foil is diverting the common mode emissions to earth.......ie, the emissions are not returning down the L or N, like they are wanted to...so this is bad...the artifical hand is a thing that worsens Common mode emissions, so that dodgy engineers like me cant pass my half-***ed kit through compliance testing.......when y caps in an smps divert emissions to the earth in the mains cable, they filter it, along with the common mode choke so its not so bad as when the y caps arent there.
,....Yes, i know all about artificial hands, as they have caught my EMC cheating hide out many a time....and unfortunately, they will continue to do so.

 

[D.A.(Tony)Stewart]

my understanding was ,,,
Conducted emission testing typically focuses on the currents flowing through its power cables. It is not typically used to measure the artificial hand current.

 

[cupoftea]

Thanks, you are correct, the artifical hand does impact the measurement of the common mode in the supplying mains cable....ayk, it worsens it.