SMPS inductor clone

Gareth Magennis

unread,

Nov 4, 2016, 5:07:24 PM11/4/16

to

OK, so I want to clone this inductor, L9.

http://elektrotanya.com/digipro_1000_sch_pcb.pdf/download.html

It is wound on a ETD29 former, I have counted the turns and the wire gauge.

What I don't know is the core material. There are several types available.

(The original has zero gap in the core)

This is a reasonably high power SMPS, running a powered speaker.
http://www.rcf.it/products/pro-speaker-systems/art-7-series/art-710-a

How would one determine the type of core used in this design?

Cheers,

Gareth.

Phil Allison

unread,

Nov 4, 2016, 11:13:51 PM11/4/16

to

Gareth Magennis wrote:
>
> OK, so I want to clone this inductor, L9.
>
> http://elektrotanya.com/digipro_1000_sch_pcb.pdf/download.html
>
>
> It is wound on a ETD29 former, I have counted the turns and the wire gauge.
>
> What I don't know is the core material. There are several types available.
>
> (The original has zero gap in the core)
>

** You sure it has no gap ???

L9 is a PFC choke and they use powdered iron cores or gapped ferrite.

Normally, the middle leg is ground short to create a gap of up to 1mm to allow DC current to pass without saturating the ferrite. Un-gapped ferrite cores are good for push pull, square wave transformers only.

You have the original core, so can measure its inductance and have a damn close look for the missing GAP.

Meanwhile, this link should supply some useful clues.

http://www.tme.eu/en/Document/d3a1e827773837de0340fb075548fa6f/etd_29_16_10.pdf

..... Phil

Gareth Magennis

unread,

Nov 5, 2016, 5:21:55 AM11/5/16

to

"Phil Allison" wrote in message
news:4adf249b-7d2b-4450...@googlegroups.com...

Thanks Phil,
I'm pretty sure it has no gap.
It's almost impossible to take these things apart without breaking the very
fragile core. I've kept all the bits so will go back and glue them together
to recheck this.
I've taken apart other SMPS transformers where the gaps are quite obvious,
even with a core in several pieces, and measurable with a micrometer without
having to reassemble.

One other thing, Digipro use enamelled stranded wire on this inductor. I
gather this has better HF performance than the much more common solid copper
enamelled wire.
A quick look for stranded enamelled wire of this guage on the Interweb
yesterday showed it is not widely available in small amounts.

Would you need to adjust the number of turns if you used solid copper
instead?

BTW, this seems to be a common failure mode of these Digipro amps - this is
the second I have seen. The windings short together, and there is no
insulation or tape applied anywhere on the windings except the final outer
wrap.

Cheers,

Gareth.

MJC

unread,

Nov 5, 2016, 6:59:12 AM11/5/16

to

In article <PwhTz.228621$9L.1...@fx26.am4>,
soundser...@outlook.com says...

>
> One other thing, Digipro use enamelled stranded wire on this inductor.
> I gather this has better HF performance than the much more common
> solid copper enamelled wire.
> A quick look for stranded enamelled wire of this guage on the Interweb
> yesterday showed it is not widely available in small amounts.

Did you search under the name "Litz wire"? (That's just something
dredged up from my electronics of decades ago.) I just got some hits,
but did not look for suppliers...

Mike.

Phil Allison

unread,

Nov 5, 2016, 7:11:51 AM11/5/16

to

Gareth Magennis wrote:
>
> >
> > OK, so I want to clone this inductor, L9.
> >
> > http://elektrotanya.com/digipro_1000_sch_pcb.pdf/download.html
> >
> > It is wound on a ETD29 former, I have counted the turns and the wire
> > gauge.
> >
> > What I don't know is the core material. There are several types
> > available.
> >
> > (The original has zero gap in the core)
> >

>
> ** You sure it has no gap ???
>
> L9 is a PFC choke and they use powdered iron cores or gapped ferrite.
>
> Normally, the middle leg is ground short to create a gap of up to 1mm to
> allow DC current to pass without saturating the ferrite. Un-gapped ferrite
> cores are good for push pull, square wave transformers only.
>
> You have the original core, so can measure its inductance and have a damn
> close look for the missing GAP.
>
> Meanwhile, this link should supply some useful clues.
>
> http://www.tme.eu/en/Document/d3a1e827773837de0340fb075548fa6f/etd_29_16_10.pdf
>

> I'm pretty sure it has no gap.
>

** There is one - it might be only a few thou of an inch, but it must be there or the whole thing would blow up.

> It's almost impossible to take these things apart without breaking the very
> fragile core. I've kept all the bits so will go back and glue them together
> to recheck this.
>
> I've taken apart other SMPS transformers where the gaps are quite obvious,
> even with a core in several pieces, and measurable with a micrometer without
> having to reassemble.
>

** This is NOT a SMPS transformer, it's a ferrite CHOKE that must pass a couple of amps of DC current. It simply HAS to have an air gap !!!

> One other thing, Digipro use enamelled stranded wire on this inductor. I
> gather this has better HF performance than the much more common solid copper
> enamelled wire.
> A quick look for stranded enamelled wire of this guage on the Interweb
> yesterday showed it is not widely available in small amounts.
>
> Would you need to adjust the number of turns if you used solid copper
> instead?

** FFS, make your own stranded wire.

Dare to be great !!!!!!

> BTW, this seems to be a common failure mode of these Digipro amps - this is
> the second I have seen. The windings short together, and there is no
> insulation or tape applied anywhere on the windings except the final outer
> wrap.
>

** Use higher temp wire, 220C stuff is readily available, the usual magnet wire is only rated at 120C at best.

Do not muck about: an active, high frequency, PFC choke is critical component.

.... Phil

Gareth Magennis

unread,

Nov 5, 2016, 11:53:41 AM11/5/16

to

"Phil Allison" wrote in message

news:148f351c-485f-42bf...@googlegroups.com...

Well there's my biggest problem, then. Unless I know what the original core
material is, I can't make one the same, the original core is in pieces.
It may not be any of the TDK formulations I can buy at Farnell.

And I guess I would need to use the same type of wire, not solid enamelled.
(or possibly rescue/repair and re-use the old winding)

Gareth.

Gareth Magennis

unread,

Nov 5, 2016, 11:57:53 AM11/5/16

to

"MJC" wrote in message news:MPG.3287cbd...@news.plus.net...

Thanks for that info, Mike.

Gareth.

Gareth Magennis

unread,

Nov 5, 2016, 6:53:12 PM11/5/16

to

"Phil Allison" wrote in message

news:148f351c-485f-42bf...@googlegroups.com...

It is becoming increasingly obvious to me that L9 in this design is ANYTHING
BUT an inductor of the appropriate inductance and the required power
handling for the job required.

It is clearly the Achilles heel of this particular design.

As such, I have given up any such silly idea as to how I might improve on
it, as I am not qualified to do so.

Thanks,

Gareth.

Look165

unread,

Nov 6, 2016, 3:52:01 AM11/6/16

to

The trace of the inductor in the schematic means with a core.

No Litz wire, it is reserved for high frequency low intensity windinds.

SMPS works at "low frequency" (some 10 kHz).

If the core is ok, just unwind and rewind with a new wire.

Gareth Magennis a écrit :