Searching \ for '[EE] Shield winding and common mode noise' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: massmind.org/techref/index.htm?key=shield+winding+common
Search entire site for: 'Shield winding and common mode noise'.

Exact match. Not showing close matches.
PICList Thread
'[EE] Shield winding and common mode noise'
2019\07\21@234422 by Manu Abraham

picon face
Hi,

I have this working SMPS with a ferrite core transformer running at
about 65kHz. It's got a flyback topology handling about 65W.

Primary is wound closest to the core and a layer of insulation exists
on top of it.
Secondary is wound on top of the Primary and a layer of insulation over it.

Improving the design, in terms of common mode noise, added a copper
foil shield winding in between the Primary and the secondary.

Stable output. Things look well indeed !



Now, moving on, decided to increase power from 65W to 150W max. To
improve coupling, the logical thought was to split the Primary into
two and the same for the secondary.

So, the logical construction goes on as follows:
Primary#1 is wound first and insulated
Secondary#1 is wound next and insulated
Primary#2 is wound next and insulated
Secondary#2 is wound last and insulated

Now, placing the shield winding has become a question, where do I
place the shield winding(s). To achieve a similar configuration as
previously, are 3 shield windings required, each between Primary and
Secondary ?

If that's true, does adding each shield winding reducing the coupling still ?

Is such a configuration advisable ?
If not, what would be a recommendable way to reduce the common mode
noise in this case where the coils are sandwiched ?

I would like to hear what folks think about it.

Best Regards,

Manu
-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2019\07\22@000812 by Bob Blick

flavicon
face
Hi Manu,
Having just dealt with a common-mode problem with an audio power amplifier, I heartily agree that an interwinding shield is a good idea.

But I really don't see how you can alternate with three or four windings. Multiple shields just don't seem like something that is going to work, partly because each shield adds to the overall diameter and pushes your windings farther and farther from the core.

Have you tried (I have not re-read the original thread) using a split bobbin method, rather than sandwiching the windings? I seem to recall you did, and it didn't work as well regarding power transfer, but it is supposed to be good regarding common-mode.

Best regards, Bob

________________________________________
From: spam_OUTpiclist-bouncesTakeThisOuTspammit.edu <.....piclist-bouncesKILLspamspam@spam@mit.edu> on behalf of Manu Abraham
Sent: Sunday, July 21, 2019 8:43 PM
To: Microcontroller discussion list - Public.
Subject: [EE] Shield winding and common mode noise

Hi,

I have this working SMPS with a ferrite core transformer running at
about 65kHz. It's got a flyback topology handling about 65W.

Primary is wound closest to the core and a layer of insulation exists
on top of it.
Secondary is wound on top of the Primary and a layer of insulation over it.

Improving the design, in terms of common mode noise, added a copper
foil shield winding in between the Primary and the secondary.

Stable output. Things look well indeed !



Now, moving on, decided to increase power from 65W to 150W max. To
improve coupling, the logical thought was to split the Primary into
two and the same for the secondary.

So, the logical construction goes on as follows:
Primary#1 is wound first and insulated
Secondary#1 is wound next and insulated
Primary#2 is wound next and insulated
Secondary#2 is wound last and insulated

Now, placing the shield winding has become a question, where do I
place the shield winding(s). To achieve a similar configuration as
previously, are 3 shield windings required, each between Primary and
Secondary ?

If that's true, does adding each shield winding reducing the coupling still ?

Is such a configuration advisable ?
If not, what would be a recommendable way to reduce the common mode
noise in this case where the coils are sandwiched ?

I would like to hear what folks think about it.

Best Regards,

Manu
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2019\07\22@012706 by Manu Abraham

picon face
Hi Bob,

I do remember the transformer problem you had previously. Having a
shield is a good idea as you say.

The shield what's in my mind is one layer of mylar insulation (less
than 66um), 1oz (66um) Cu foil and another layer of mylar insulation.
The whole thing adds to 0.2mm thickness. An additional 2x shield foil
means 0.4mm lost in the winding area.

In the preliminary situation, the idea was to have the entire primary
winding in one go and then the secondary later, but got a suggestion
that the coupling would get a hit and was suggested to split the
windings to sandwich them, primary-secondary-primary-secondary.

Tried this topology, it works well, but the problem of common mode
noises arises. The noise does severely affect equipment powered from
it. Thus the shielding thought came to my mind.

Multiple shields, other than loosing precious winding area, will it
affect magnetic coupling in an adverse way ?

Having a split bobbin does not work in acceptable manner, as the
magnetic coupling is severely affected in which case. Commom mode
chokes, EHT transformers possible though.In the very first case the
windings had to be sandwiched to improve coupling ..

Thanks,

Manu

On Mon, Jul 22, 2019 at 9:39 AM Bob Blick <bobblickspamKILLspamoutlook.com> wrote:
{Quote hidden}

-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2019\07\22@042212 by AB Pearce - UKRI STFC

face picon face
The major problem I see with having split windings is that you will increase the winding capacitance to ground which will probably not be desirable at the frequencies you are proposing. The shield is supposed to have a connection brought out and connected to ground if you are seriously looking at minimising noise coupling between primary and secondary, and having such a capacitance from the drive end of the winding (the end connected to the transistor collector or drain) is likely to produce undesirable effects.

Personally I would wind it with a single shield and then make sure that on the primary side the winding end against the shield is connected to the supply rather than the transistor.



{Original Message removed}

2019\07\22@054621 by Manu Abraham

picon face
On Mon, Jul 22, 2019 at 1:53 PM AB Pearce - UKRI STFC
<alan.b.pearcespamspam_OUTstfc.ac.uk> wrote:
>
> The major problem I see with having split windings is that you will increase the winding capacitance to ground which will probably not be desirable at the frequencies you are proposing. The shield is supposed to have a connection brought out and connected to ground if you are seriously looking at minimising noise coupling between primary and secondary, and having such a capacitance from the drive end of the winding (the end connected to the transistor collector or drain) is likely to produce undesirable effects.
>


One thought to reduce winding capacitance, would be bank winding. Does
that solve the capacitance issue ?
As in (http://tiny.cc/iq519y) ?
In the original case, bank winding was used. The same possibly can be
applied, if that helps.



> Personally I would wind it with a single shield and then make sure that on the primary side the winding end against the shield is connected to the supply rather than the transistor.
>
>

I was of the assumption that the shield should be Grounded. But you
mention Supply. I've been looking at different designs. I see that
most of them have the shield connected to the Aux/Bias power supply
(not ground). Any idea, why ?

-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2019\07\22@095442 by AB Pearce - UKRI STFC

face picon face
> Personally I would wind it with a single shield and then make sure that on the primary side the winding end against the shield is connected to the supply rather than the transistor.
>
>> I was of the assumption that the shield should be Grounded. But you mention Supply.
>> I've been looking at different designs. I see that most of them have the shield connected
>> to the Aux/Bias power supply (not ground). Any idea, why ?

The shield should be grounded, but here I was talking about the supply end of the primary winding.

Connecting to some other supply may allow lower capacitance or leakage or coupling between windings. I don't recall seeing a circuit done this way though.
Perhaps a link to a reference is in order?



-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2019\07\22@110715 by Manu Abraham

picon face
On Mon, Jul 22, 2019 at 7:27 PM AB Pearce - UKRI STFC
<@spam@alan.b.pearceKILLspamspamstfc.ac.uk> wrote:
>
> > Personally I would wind it with a single shield and then make sure that on the primary side the winding end against the shield is connected to the supply rather than the transistor.


Slightly confused by this statement.

If you look at the presentation from Ridley, he states the power
supply end needs to be the farthest from output ground. (Page #7) for
a low noise configuration.

http://tiny.cc/pgm29y

In which case a bank winding would allow the most distance between the
2 ends, right ?

{Quote hidden}

Sure, I was looking at this one, atm:

http://www.ti.com/lit/ug/sluuax9b/sluuax9b.pdf
(Page #5, #20)
-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2019\07\22@125510 by Manu Abraham

picon face
On Mon, Jul 22, 2019 at 8:36 PM Manu Abraham <KILLspamabraham.manuKILLspamspamgmail.com> wrote:
{Quote hidden}

Additionally, tripped up on this video on youtube with a similar
configuration. That was indeed a coincidence.

http://tiny.cc/0xv29y

2 shields, both connected to the Aux winding. Strange though, why Aux winding ..
-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

More... (looser matching)
- Last day of these posts
- In 2019 , 2020 only
- Today
- New search...