Using a dual-inductor line filter as a DC-DC transformer?

I have seen a design for a high-power (200kW) AC motor controller that uses 
dual-inductor AC line chokes for a DC-DC converter for IGBT gate drive power 
supplies.
http://enginuitysystems.com/pix/electronics/200kW_AC_Controller_p3.png

I have found some chokes rated 500 uH and 1 amp, with 250 VAC rated 
isolation, for less than a dollar each:
http://www.mouser.com/ds/2/212/KEM_LF0016_SU9V-9H-372621.pdf
http://www.mouser.com/ProductDetail/KEMET-NEC-Tokin/SU9H-10005/?qs=sGAEpiMZZMsVJzu5wKIZCRKkZmKdbMQOj5JKMlEVxjU%3d

I'm going to order a few of these to see if they will work for the SCR 
trigger board I am designing. I might not trust them for 480 VAC mains, but 
there are some applications that will be used on 120 or 240 VAC mains which 
should be safe. The graph for this part shows an impedance of about 300 ohms 
at 60 kHz, so I would figure this works out to a magnetizing current of 
about 40 mA for a 12V square wave. The split bobbin probably makes for a 
fair amount of leakage inductance, but my prototype of similar construction 
has about 5% (25 uH) and they work just fine.

Has anyone tried this? Are there any issues with such a design?

Thanks,

paul 

CMCs are only good for about 1-2W.  If that's all you need, then go for it.

Hardly any need for current limiting either, thanks to the huge leakage. 
Just make the driver a little oversized.  It's "impedance protected" like 
all those tiny AC motors!  ;-)

Wowww, that schematic is laughable.  ATtiny instead of a 555.  Driving an 
IXDD604.  Guessing they don't know their asses from their stack of 
cookbooks.  Wait, the fine print says "from Instructables"?  The closer you 
look at it, the harder you laugh...LOL!

Tim

-- 
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Contract Design
Website: http://seventransistorlabs.com

"P E Schoen" <paul@pstech-inc.com> wrote in message 
news:nratg7$u36$1@dont-email.me...
>I have seen a design for a high-power (200kW) AC motor controller that uses 
>dual-inductor AC line chokes for a DC-DC converter for IGBT gate drive 
>power supplies.
> http://enginuitysystems.com/pix/electronics/200kW_AC_Controller_p3.png
>
> I have found some chokes rated 500 uH and 1 amp, with 250 VAC rated 
> isolation, for less than a dollar each:
> http://www.mouser.com/ds/2/212/KEM_LF0016_SU9V-9H-372621.pdf
> http://www.mouser.com/ProductDetail/KEMET-NEC-Tokin/SU9H-10005/?qs=sGAEpiMZZMsVJzu5wKIZCRKkZmKdbMQOj5JKMlEVxjU%3d
>
> I'm going to order a few of these to see if they will work for the SCR 
> trigger board I am designing. I might not trust them for 480 VAC mains, 
> but there are some applications that will be used on 120 or 240 VAC mains 
> which should be safe. The graph for this part shows an impedance of about 
> 300 ohms at 60 kHz, so I would figure this works out to a magnetizing 
> current of about 40 mA for a 12V square wave. The split bobbin probably 
> makes for a fair amount of leakage inductance, but my prototype of similar 
> construction has about 5% (25 uH) and they work just fine.
>
> Has anyone tried this? Are there any issues with such a design?
>
> Thanks,
>
> paul 

On Wed, 14 Sep 2016 03:16:23 -0400, "P E Schoen" <paul@pstech-inc.com>
wrote:

>I have seen a design for a high-power (200kW) AC motor controller that uses 
>dual-inductor AC line chokes for a DC-DC converter for IGBT gate drive power 
>supplies.
>http://enginuitysystems.com/pix/electronics/200kW_AC_Controller_p3.png
>
>I have found some chokes rated 500 uH and 1 amp, with 250 VAC rated 
>isolation, for less than a dollar each:
>http://www.mouser.com/ds/2/212/KEM_LF0016_SU9V-9H-372621.pdf
>http://www.mouser.com/ProductDetail/KEMET-NEC-Tokin/SU9H-10005/?qs=sGAEpiMZZMsVJzu5wKIZCRKkZmKdbMQOj5JKMlEVxjU%3d
>
>I'm going to order a few of these to see if they will work for the SCR 
>trigger board I am designing. I might not trust them for 480 VAC mains, but 
>there are some applications that will be used on 120 or 240 VAC mains which 
>should be safe. The graph for this part shows an impedance of about 300 ohms 
>at 60 kHz, so I would figure this works out to a magnetizing current of 
>about 40 mA for a 12V square wave. The split bobbin probably makes for a 
>fair amount of leakage inductance, but my prototype of similar construction 
>has about 5% (25 uH) and they work just fine.
>
>Has anyone tried this? Are there any issues with such a design?
>
>Thanks,
>
>paul 

You could buy a complete pottted dc/dc converter for a few dollars.
There are also reasonably priced converters rated for kilovolts of
isolation.

It's fun to design stuff, but dc/dc bricks are cheaper than you can
buy the parts.




-- 

John Larkin         Highland Technology, Inc

lunatic fringe electronics 

John Larkin wrote...
>
> You could buy a complete potted dc/dc converter for a few dollars.
> There are also reasonably priced converters rated for kilovolts of
> isolation.

 I like the +20,-4V 100mA modules from MornSun (QA01C)
 and MPD (IG120-20).  3.5kVac.  Same pinout.  Only
 3.5pF of I/O capacitance!  Nice.  About $6 qty 10.


-- 
 Thanks,
    - Win

On 14 Sep 2016 07:46:15 -0700, Winfield Hill
<hill@rowland.harvard.edu> wrote:

>John Larkin wrote...
>>
>> You could buy a complete potted dc/dc converter for a few dollars.
>> There are also reasonably priced converters rated for kilovolts of
>> isolation.
>
> I like the +20,-4V 100mA modules from MornSun (QA01C)
> and MPD (IG120-20).  3.5kVac.  Same pinout.  Only
> 3.5pF of I/O capacitance!  Nice.  About $6 qty 10.

That's nice, two high-isolation supplies in one brick. Just right for
powering opamps, too. Sometimes just a little negative voltage is all
you need.

I use the single and dual CUI bricks even when I don't need isolation.
This one is slick:

www.cui.com/product/resource/pds1-m.pdf



-- 

John Larkin         Highland Technology, Inc

lunatic fringe electronics 

"Tim Williams"  wrote in message news:nrb5f4$l0f$1@dont-email.me...

> CMCs are only good for about 1-2W.  If that's all you need, then go for 
> it.

Actual gate current about 200-300 mA with 12 VDC open circuit. So maybe 
2.4-3.6 watts, but if the impedance causes the output voltage to drop 
(without much power due to inductance) then it might be fine.

> Hardly any need for current limiting either, thanks to the huge leakage. 
> Just make the driver a little oversized.  It's "impedance protected" like 
> all those tiny AC motors!  ;-)

> Wowww, that schematic is laughable.  ATtiny instead of a 555.  Driving an 
> IXDD604.  Guessing they don't know their asses from their stack of 
> cookbooks.  Wait, the fine print says "from Instructables"?  The closer 
> you look at it, the harder you laugh...LOL!

That is actually an earlier version of the AC controller sold by 
PaulAndSabrina, for DIY EVs, and apparently they work quite well:
http://www.paulandsabrinasevstuff.com/onlinestore.html#!/EV-Boards/c/16287307/offset=0&sort=normal

The ATtiny (or an equivalent Microchip PIC) is really a much better choice 
than a 555 because you can program the dead time between alternate square 
wave pulses. The cost is comparable, and rather insignificant. The IXDN604, 
with 4A output and only about $1.80, is also a pretty good alternative to a 
four-MOSFET bridge.

Paul

"John Larkin"  wrote in message 
news:memitb1oel0hkvi1ls90e3g7peerppfstq@4ax.com...

> You could buy a complete pottted dc/dc converter for a few dollars. There 
> are also reasonably priced converters rated for kilovolts of isolation.

> It's fun to design stuff, but dc/dc bricks are cheaper than you can buy 
> the parts.

The 480 VAC mains presents a special problem, which requires a "reinforced" 
isolation having a physical barrier between the windings capable of safe 
operation. Here is more information:

http://www.richardsonrfpd.com/resources/RellDocuments/SYS_29/RECOM_High_Isolation_DC_DC_Converters_and_Applications.pdf

Previously, the only DC-DC converters I found with such specifications were 
these, which are $40 or more:

http://www.mouser.com/ProductDetail/TRACO-Power/THB-3-1211/?qs=sGAEpiMZZMvGsmoEFRKS8GuhBnqedIdiEgkUN%252bAArjjvLaJ4QOcihA%3d%3d

http://www.mouser.com/ds/2/687/thb3-519372.pdf

or

http://www.mouser.com/ProductDetail/TRACO-Power/THP-3-2412/?qs=sGAEpiMZZMvGsmoEFRKS8GuhBnqedIdiJExaWj7K%252bxSiqfkOOWH72w%3d%3d

http://www.mouser.com/ds/2/687/thp3-537683.pdf

Note that only the first one is rated at 4800 VAC and suitable for up to 
1000 VRMS working voltage. A 3000 VAC rating is supposedly only rated to 300 
VRMS continuous working voltage. That unit is 5VDC output at 600 mA, which 
should be fine for SCR gates that typically are about 1 volt at that 
current.

For IGBTs, SIC, and MOSFETs, Murata has some good modules for about $24 
each:

http://www.mouser.com/ProductDetail/Murata-Power-Solutions/MGJ6T12150505MC-R7/?qs=sGAEpiMZZMvGsmoEFRKS8NEjhuwmH457h5W%252bhzK9PoBVp6V7Jg2uuA%3d%3d

http://www.mouser.com/ds/2/281/kdc_mgj6-844919.pdf

Paul 

"John Larkin"  wrote in message 
news:4lritbhm1beapqdvvgg29vr0vakpumoa45@4ax.com...

> On 14 Sep 2016 07:46:15 -0700, Winfield Hill <hill@rowland.harvard.edu> 
> wrote:

>> I like the +20,-4V 100mA modules from MornSun (QA01C) and MPD (IG120-20). 
>> 3.5kVac.  Same pinout.  Only 3.5pF of I/O capacitance!  Nice.  About $6 
>> qty 10.

> That's nice, two high-isolation supplies in one brick. Just right for 
> powering opamps, too. Sometimes just a little negative voltage is all you 
> need.

> I use the single and dual CUI bricks even when I don't need isolation.
> This one is slick:

> www.cui.com/product/resource/pds1-m.pdf

Murata also makes some nice 2W DC-DC converters for IGBTs, MOSFETs, and SiC 
devices (about $8):

http://www.mouser.com/ProductDetail/Murata-Power-Solutions/MGJ2D121509SC/?qs=sGAEpiMZZMvGsmoEFRKS8O6J06AyJhRctng5IFEResaltu7JfAQRSg%3d%3d

http://www.mouser.com/ds/2/281/kdc_mgj2-318966.pdf

Paul


-- 

John Larkin         Highland Technology, Inc

lunatic fringe electronics 

P E Schoen wrote...
>
> "John Larkin"  wrote ...
>> Winfield Hill wrote:
>
>>> I like the +20,-4V 100mA modules from MornSun (QA01C) and MPD (IG120-20). 
>>> 3.5kVac.  Same pinout.  Only 3.5pF of I/O capacitance!  Nice.  About $6 
>>> qty 10.
>
>> That's nice, two high-isolation supplies in one brick. Just right for 
>> powering opamps, too. Sometimes just a little negative voltage is all you 
>> need.
>
>> I use the single and dual CUI bricks even when I don't need isolation.
>> This one is slick:  www.cui.com/product/resource/pds1-m.pdf
>
> Murata also makes some nice 2W DC-DC converters for IGBTs,
> MOSFETs, and SiC devices (about $8):
> http://www.mouser.com/ds/2/281/kdc_mgj2-318966.pdf

 Lots of choices.  Quote: "Ultra low coupling capacitance",
 but they don't give a value.  3.5pF is sweet spec value.


-- 
 Thanks,
    - Win

On Wednesday, September 14, 2016 at 3:16:31 AM UTC-4, P E Schoen wrote:
> I have seen a design for a high-power (200kW) AC motor controller that uses 
> dual-inductor AC line chokes for a DC-DC converter for IGBT gate drive power 
> supplies.
> http://enginuitysystems.com/pix/electronics/200kW_AC_Controller_p3.png
> 
> I have found some chokes rated 500 uH and 1 amp, with 250 VAC rated 
> isolation, for less than a dollar each:
> http://www.mouser.com/ds/2/212/KEM_LF0016_SU9V-9H-372621.pdf
> http://www.mouser.com/ProductDetail/KEMET-NEC-Tokin/SU9H-10005/?qs=sGAEpiMZZMsVJzu5wKIZCRKkZmKdbMQOj5JKMlEVxjU%3d
> 
> I'm going to order a few of these to see if they will work for the SCR 
> trigger board I am designing. I might not trust them for 480 VAC mains, but 
> there are some applications that will be used on 120 or 240 VAC mains which 
> should be safe. The graph for this part shows an impedance of about 300 ohms 
> at 60 kHz, so I would figure this works out to a magnetizing current of 
> about 40 mA for a 12V square wave. The split bobbin probably makes for a 
> fair amount of leakage inductance, but my prototype of similar construction 
> has about 5% (25 uH) and they work just fine.
> 
> Has anyone tried this? Are there any issues with such a design?

I think you'll find the choke cores don't have much of a volt-time product because in the typical application the differential voltage induces zero core flux and the common mode operation is at relatively small voltages. Another problem is they should be using core material formulated to be lossy at high frequency- similar to suppression beads.

> 
> Thanks,
> 
> paul