Searching \ for '[EE] Inductor power rating for LTC3112' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: techref.massmind.org/techref/power.htm?key=power
Search entire site for: 'Inductor power rating for LTC3112'.

Exact match. Not showing close matches.
PICList Thread
'[EE] Inductor power rating for LTC3112'
2011\01\13@212505 by V G

picon face
Hey all,

So I looked through the datasheet for the LTC3112 chip (
http://cds.linear.com/docs/Datasheet/3112f.pdf) and went to my local
electronics store (for those interested, the one on College st near U of T)..
I bought a 4.7 uH inductor but they only carried those with 1/2 watt rating..
I bought it anyway.

I'm wondering if the power rating on this inductor is sufficient for this
chip supplying 1 Amp at 5 Volts from a 4.2 - 5.8 Volt supply.

I looked through the datasheet and the section on selecting an inductor
didn't comment on the power rating

2011\01\13@233624 by IVP

face picon face
> I'm wondering if the power rating on this inductor is sufficient
> for this chip supplying 1 Amp at 5 Volts from a 4.2 - 5.8 Volt
> supply

Page 26 shows a current wave for that particular example, 5V/2A,
of about 1A peak. The demo overlay on Page 17 gives an idea of
physical size. For this type of IC they do not use the big current-
carrying toroidals, as the internals have been designed to avoid both
the size and expense. On other switcher ICs I've used physically
small bobbin inductors in the 3u3H to 15uH range. Some datasheets
have provided part numbers (eg Toko) that I could cross-reference
to locally available one

2011\01\13@233707 by Kerry Wentworth

flavicon
face
Inductors are normally rated by current, not power.  Calculate max inductor current, and see if  I^2 * DC resistance of the inductor is less than 1/2 watt.  If it is, you will probably be OK.  That assumes that the inductor won't saturate at less than it's rated dissipation.

Kerry



V G wrote:
{Quote hidden}

-- Internal Virus Database is out-of-date.
Checked by AVG Anti-Virus.
Version: 7.0.289 / Virus Database: 267.11.13 - Release Date: 10/6/05

2011\01\14@032454 by Michael Watterson

face picon face
On 14/01/2011 02:24, V G wrote:
> Hey all,
>
> So I looked through the datasheet for the LTC3112 chip (
> http://cds.linear.com/docs/Datasheet/3112f.pdf) and went to my local
> electronics store (for those interested, the one on College st near U of T).
> I bought a 4.7 uH inductor but they only carried those with 1/2 watt rating.
> I bought it anyway.
>
> I'm wondering if the power rating on this inductor is sufficient for this
> chip supplying 1 Amp at 5 Volts from a 4.2 - 5.8 Volt supply.
>
> I looked through the datasheet and the section on selecting an inductor
> didn't comment on the power rating.
Power rating depends on frequency and waveform and DC component.
Current rating depends on wire and core frequency and waveform and DC component.

If it's not air core you need the correct ferrite or dust iron core for the frequency/Application

There are many ferrite types

2011\01\14@033228 by RussellMc

face picon face
part 1 981 bytes content-type:text/plain; charset="iso-8859-1" (decoded quoted-printable)

> So I looked through the datasheet for the LTC3112 chip (
> http://cds.linear.com/docs/Datasheet/3112f.pdf) and went to my local
> electronics store (for those interested, the one on College st near U of T).
> I bought a 4.7 uH inductor but they only carried those with 1/2 watt rating.
> I bought it anyway.
>
> I'm wondering if the power rating on this inductor is sufficient for this
> chip supplying 1 Amp at 5 Volts from a 4.2 - 5.8 Volt supply.

See Kerry's comment on current rating
Inductor must take peak current carried without saturating.
For "saturation" see gargoyle.
At a miniumm you'd usually want Isat > Imax and 4 x ma be good. In
certain topologies (which you have not got) it may be much more.
Look here 9in Chinese) for SF73 series which are larger than you need.
data sheet attached - if you see this it fitted inside the file size limit.


                Russell McMahon


part 2 31763 bytes content-type:application/pdf; name="DS INDUCTOR SF73 Fenfa.pdf" (decode)

part 3 181 bytes content-type:text/plain; name="ATT00001.txt"
(decoded base64)

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

2011\01\14@062509 by V G

picon face
I ordered some samples of inductors from this site:
http://www.coilcraft.com/

But after reading some of your replies, I'm not so sure they're the right
type. I'm not sure which type exactly I need. There seem to be so many.

Also, the 1/2 watt inductor I bought had a resistance rating on there, so I
think safely assuming max current is 1 A, (1^2 * 0.2 = 0.2 W), so I *should*
be safe

2011\01\14@070559 by Kerry Wentworth

flavicon
face
Inductor current is not the same as output current.  You need to calculate the inductor current.  But I don't think it would be unreasonable to build one and see what happens, if you like.

Kerry


V G wrote:
{Quote hidden}

-- Internal Virus Database is out-of-date.
Checked by AVG Anti-Virus.
Version: 7.0.289 / Virus Database: 267.11.13 - Release Date: 10/6/05

2011\01\14@071752 by V G

picon face
On Fri, Jan 14, 2011 at 7:06 AM, Kerry Wentworth <
spam_OUTkwentworthTakeThisOuTspamskunkworksnh.com> wrote:

> Inductor current is not the same as output current.  You need to
> calculate the inductor current.  But I don't think it would be
> unreasonable to build one and see what happens, if you like.


I was thinking of that, but I have no idea of all the properties and
equations related to inductors. I've never used them before. Might as well
just stick with off the shelf components

2011\01\14@073652 by Kerry Wentworth

flavicon
face
They give a formula in the spec sheet (page 14), I'm just too lazy to crunch the numbers.  A quick check indicates the current will be in the neighborhood of 1/4 amp, so you should be fine.

Kerry


V G wrote:
{Quote hidden}

-- Internal Virus Database is out-of-date.
Checked by AVG Anti-Virus.
Version: 7.0.289 / Virus Database: 267.11.13 - Release Date: 10/6/05

2011\01\14@073831 by RussellMc

face picon face
> I ordered some samples of inductors from this site:

       http://www.coilcraft.com/

> But after reading some of your replies, I'm not so sure they're the right
> type.

Yes. We have much the same problem :-) - as you didn't say which type
you ordered.

> I'm not sure which type exactly I need. There seem to be so many.

How did they rate wrt:

       >> At a miniumm you'd usually want Isat > Imax and 4 x ma be good.

and how did this inform your understanding

       >> Inductor must take peak current carried without saturating.
       >> For "saturation" see gargoyle.

Especially the stone creature part?



2011\01\14@074536 by RussellMc

face picon face
> I ordered some samples of inductors from this site:
> http://www.coilcraft.com/
>
> But after reading some of your replies, I'm not so sure they're the right
> type.

Did you try usig any of

  http://www.coilcraft.com/design_tools.cfm


?

2011\01\14@075520 by Olin Lathrop

face picon face
V G wrote:
> I bought a 4.7 uH inductor but they only carried those with
> 1/2 watt rating. I bought it anyway.
>
> I'm wondering if the power rating on this inductor is sufficient for
> this chip supplying 1 Amp at 5 Volts from a 4.2 - 5.8 Volt supply.

It's impossible to tell based on just the inductance and the power rating of
the inductor.  The important parameter of the inductor is its saturation
current.  That's the current at which the inductance drastically plummets
for higher current values.  In other words, it stops being a inductor and
looks primarily like a resistor, which has the value of it series
resistance.

The power rating comes from how much heat it can dissipate.  Ideal inductors
can't dissipate power, so all the heat comes from the equivalent series
resistance.  If you know what that is, you can calculate the RMS current
that will produce the specified power dissipation.

The fact that your inductor came with a power rating instead of a saturation
current rating and the fact that it's only rated for 1/2 Watt makes me think
it's a RF choke.  Those are meant for RF applications where the impedences
are generally high and a few Ohms resistance in the inductor not a big deal..
For power applications the frequencies are lower, the currents high, the
impedences therefore lower, and the inductor resistance a important factor.

Coilcraft is pretty good about free samples in small quantities.  They make
all kinds of power inductors and specify them well.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000

2011\01\14@105755 by William \Chops\ Westfield

face picon face

On Jan 14, 2011, at 12:31 AM, RussellMc wrote:

> Inductor must take peak current carried without saturating.

What happens if you build an otherwise correct SMPS but your inductor  saturation current is too low?  Is this easy to recognize from steady- state symptoms, or does your expensive SMPS chip mysteriously fail...

BillW

2011\01\14@110732 by Spehro Pefhany

picon face
At 10:57 AM 14/01/2011, you wrote:

>On Jan 14, 2011, at 12:31 AM, RussellMc wrote:
>
> > Inductor must take peak current carried without saturating.
>
>What happens if you build an otherwise correct SMPS but your inductor
>saturation current is too low?  Is this easy to recognize from steady-
>state symptoms, or does your expensive SMPS chip mysteriously fail...
>
>BillW

A well-designed _current-mode_ topology circuit should not fail,
but the inductor will get very hot and you won't be able to get
the desired output.

>Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
speffspamKILLspaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

2011\01\14@112503 by Olin Lathrop

face picon face
'William Chops" Westfield ' <.....westfwKILLspamspam.....mac.com wrote:
> What happens if you build an otherwise correct SMPS but your inductor
> saturation current is too low?

Various things are possible.  The most obvious symptoms are usually
considerable loss of efficiency and excessive heating the switching
transistor.  The inductor may also get rather warmer than expected.

The switchover from inductor to low value resistor operation can also cause
transients that can cause problems in other parts of the circuit.

> Is this easy to recognize from steady-
> state symptoms, or does your expensive SMPS chip mysteriously fail...

Often the output voltage of the power supply is still about right, so
downstream things generally don't fail.  The switching transistor takes a
beating, so it is the most likely thing to fail.  If it's inside the SMPS
chip, then that chip may fail.

We had a case recently of a SMD inductor getting installed rotated 90
degrees.  That essentially replaced the inductor with a dead short.  The
output voltage was still right and the rest of the board worked.  In that
case the switch was internal to the SMPS chip.  It got hot, but didn't fry.
Everything was fine when the inductor was removed and mounted correctly.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000

2011\01\14@224641 by V G

picon face
On Fri, Jan 14, 2011 at 7:56 AM, Olin Lathrop <EraseMEolin_piclistspam_OUTspamTakeThisOuTembedinc.com>wrote:

> The fact that your inductor came with a power rating instead of a
> saturation
> current rating and the fact that it's only rated for 1/2 Watt makes me
> think
> it's a RF choke.  Those are meant for RF applications where the impedences
> are generally high and a few Ohms resistance in the inductor not a big
> deal.
> For power applications the frequencies are lower, the currents high, the
> impedences therefore lower, and the inductor resistance a important factor.
>
> Coilcraft is pretty good about free samples in small quantities.  They make
> all kinds of power inductors and specify them well.


Thanks. I'm going to give this a try anyway and test out the performance
under various loads, including a short

2011\01\14@231827 by RussellMc

face picon face
> We had a case recently of a SMD inductor getting installed rotated 90
> degrees.  That essentially replaced the inductor with a dead short.  The
> output voltage was still right and the rest of the board worked.  In that
> case the switch was internal to the SMPS chip.  It got hot, but didn't fry.
> Everything was fine when the inductor was removed and mounted correctly.

That was, presumably, a buck converter. Which would have become a
series linear regulator.

Do that with a boost converter and YMWV.


            Russell

2011\01\17@070813 by V G

picon face
On Thu, Jan 13, 2011 at 11:35 PM, IVP <joecolquittspamspam_OUTclear.net.nz> wrote:

> Page 26 shows a current wave for that particular example, 5V/2A,
> of about 1A peak.


Hi!

* I've been looking at that page. Where is the zero line on that graph for
the current? I see a difference in current of about 1 A, but what is the
*absolute* max current?
* The plastic bag that the inductor came with says that it's a "power"
inductor rated at 1/2 W with a resistance of 0.3 ohms. I then measured the
resistance with my multimeter and found it to be 0.7 ohms, but that could be
wrong?
* At 0.3 ohms, (I^2)*R < 0.5 W, I would get around 1.29 A max. What are the
effects of pushing more than 1.29 A through this thing?
* I don't plan on drawing more than 1 A, so the peak current through the
inductor should be less than the waveform shown on page 26, and I should be
good to go

2011\01\17@073711 by alan.b.pearce

face picon face
> > Page 26 shows a current wave for that particular example, 5V/2A,
> > of about 1A peak.
>
>
> Hi!
>
> * I've been looking at that page. Where is the zero line on that graph for
> the current? I see a difference in current of about 1 A, but what is the
> *absolute* max current?
> * The plastic bag that the inductor came with says that it's a "power"
> inductor rated at 1/2 W with a resistance of 0.3 ohms. I then measured the
> resistance with my multimeter and found it to be 0.7 ohms, but that could be
> wrong?
> * At 0.3 ohms, (I^2)*R < 0.5 W, I would get around 1.29 A max. What are the
> effects of pushing more than 1.29 A through this thing?
> * I don't plan on drawing more than 1 A, so the peak current through the
> inductor should be less than the waveform shown on page 26, and I should be
> good to go?

I think what you need to do is look at some application notes on SMPS design. Linear technology have some good ones on how to specify inductors.

You may find appendix B of AN35 useful.
AN19 LT1070 family design manual
AN44 page 17 onwards.

I am not familiar with the LTC3112, but does its web page reference any app notes on circuit design?


-- Scanned by iCritical.

2011\01\17@100834 by Oli Glaser

flavicon
face
On 17/01/2011 12:07, V G wrote:
> On Thu, Jan 13, 2011 at 11:35 PM, IVP<@spam@joecolquittKILLspamspamclear.net.nz>  wrote:
>
>> Page 26 shows a current wave for that particular example, 5V/2A,
>> of about 1A peak.
>
> Hi!
>
> * I've been looking at that page. Where is the zero line on that graph for
> the current? I see a difference in current of about 1 A, but what is the
> *absolute* max current?

That page is showing waveforms at a particular input voltage (of around 12V I think) so you can't really go from that unless you are doing exactly the same. At different input voltages/loads things will change.

> * The plastic bag that the inductor came with says that it's a "power"
> inductor rated at 1/2 W with a resistance of 0.3 ohms. I then measured the
> resistance with my multimeter and found it to be 0.7 ohms, but that could be
> wrong?
> * At 0.3 ohms, (I^2)*R<  0.5 W, I would get around 1.29 A max. What are the
> effects of pushing more than 1.29 A through this thing?
> * I don't plan on drawing more than 1 A, so the peak current through the
> inductor should be less than the waveform shown on page 26, and I should be
> good to go?

Not necessarily (I think quite likely not)
I can remember whether you are using this for buck/boost but I agree with Alan that looking at some app notes would be a good idea. Also I really recommend getting LTSpice if you have not already (great free simulator in general) and using their example setups to simulate.
I just looked and they don't have the 3112 but they do have an example of the 3533 which is quite similar (in <install folder>/examples/jigs) - I think a lot would be gained from running this example and changing various parameters (input voltage, inductor series resistance etc) and watching the pulse width, inductor current, output current etc. For example what happens to the current when you have the in voltage lower than the out voltage and vice versa? What happens to the pulse width when in voltage is higher? What happens when you raise the  inductor series resistance? What happens when you change the load?
This plus going over the theory thoroughly should help I think (i.e. don't just take the simulation results completely for granted - not always correct/comparable to real world conditions - e.g. sometimes totally wrong/misleading if set up the wrong way, although with the LT examples they *should* be set up okay at least to start with)

Also, doing similar in the "real world" will be good learning too (i.e. just shove the inductor in and see what happens.. :-) )



2011\01\17@103425 by RussellMc

face picon face
>> * I don't plan on drawing more than 1 A, so the peak current through the
>> inductor should be less than the waveform shown on page 26, and I should be
>> good to go?

Reading what people have said so far and doing what's suggested would
probably be helpful.
Do you understand core saturation in general terms (or better)?



2011\01\17@142148 by Olin Lathrop

face picon face
V G wrote:
> * The plastic bag that the inductor came with says that it's a "power"
> inductor rated at 1/2 W with a resistance of 0.3 ohms. I then
> measured the resistance with my multimeter and found it to be 0.7
> ohms, but that could be wrong?

Yes, if the inductor is of questionable origin or from a questional
supplier.  More likely though is that your ohm meter isn't that good below 1
ohm.  To check, short the leads together and see what it says.  Maybe it as
a option to zero it with the leads shorted.

> * At 0.3 ohms, (I^2)*R < 0.5 W, I would get around 1.29 A max.

Yes, that seems to add up.

> What
> are the effects of pushing more than 1.29 A through this thing?

It will get hotter than it's supposed to.  At first it will probably just
not live up to datasheet specs.  If the overheat is significant, it will
fail.

> * I don't plan on drawing more than 1 A, so the peak current through
> the inductor should be less than the waveform shown on page 26, and I
> should be good to go?

I don't know what's on page 26, but you've already shown that 1.29A is safe
in terms of total dissipation.  That tells you the upper limit of the RMS
current.  The other parameter you need is the saturation current.  That is
the instantaneous upper current limit for most normal circuits where the
inductor is expected to remain acting like a inductor.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000

2011\01\18@012702 by V G

picon face
On Mon, Jan 17, 2011 at 10:08 AM, Oli Glaser <KILLspamoli.glaserKILLspamspamtalktalk.net>wrote:

{Quote hidden}

Ok, I just ran the simulation. How do I get it to simulate more than 2 ms?

This is what it looks like:
http://img835.imageshack.us/img835/879/capturebx.png

Seems that the current through the inductor is a lot higher than 2.7 A

2011\01\18@013026 by V G

picon face
On Tue, Jan 18, 2011 at 1:26 AM, V G <RemoveMEx.solarwind.xTakeThisOuTspamgmail.com> wrote:

{Quote hidden}

Spoke too soon. I changed the simulation time and this is what I got:
http://img718.imageshack.us/img718/5750/captureqwp.pn

2011\01\18@020250 by V G

picon face
1. The SPICE simulation is proving to be very useful and interesting. By
changing the parameters, I can see that the current through the inductor
depends heavily on the load. So at the worst case, I can tell the iPhone to
draw 500 mA instead of 1 A to charge which will "for sure" keep me in a safe
range. Also, current through the inductor increases as the input voltage
decreases. So at a minimum of 4 V input, it seems that I'm getting around
1.35 A through the inductor at a 1 A load (5 V output). At a "normal" input
voltage of 4.8 V, current through the inductor is around 1.1 A, which seems
to be safe.

I haven't messed around with all the parameters so I could be wrong, but I
just changed some values around to better fit my scenario and the above is
what I've observed.

2. My multimeter shows 0.0 ohms when shorted, so it's properly zeroed as far
as I know. Hopefully the inductor that I bought does have a lower series
resistance than what was labelled. I can put some thermal compound and a
heatsink on the PCB to be safe. This is all assuming that it doesn't
"saturate" and the power supply can properly source 1 A

2011\01\18@022640 by Oli Glaser

flavicon
face
On 18/01/2011 06:30, V G wrote:
>> >
>> >  Ok, I just ran the simulation. How do I get it to simulate more than 2 ms?
>> >
>> >  This is what it looks like:
>> >  img835.imageshack.us/img835/879/capturebx.png
>> >
>> >  Seems that the current through the inductor is a lot higher than 2.7 A.
>> >
> Spoke too soon. I changed the simulation time and this is what I got:
> http://img718.imageshack.us/img718/5750/captureqwp.png

Okay - if you haven't used LTSpice before then take a little time to read the help file (how to set the simulation up, the waveform viewer, spice commands and so on)
Couple of useful tips:
You can zoom in on the waveforms by dragging your mouse around the area of interest (all detailed in the waveform viewer section) so do this to see exactly what the various parts are doing. Also you can set the simulation to run for say 3ms but only display from 1ms onward (time to start saving data)
Holding Alt whilst clicking on component will provide power waveform (cursor changes to thermometer)

After reading, getting the hang of it a bit, set it up as near to your proposed circuit as possible e.g. set input voltage, divider ratio (to set Vout), adjust inductor series resistance to match what the one you bought, set load to roughly simulate what you will be attaching (IIRC it was going to draw <1A, also you can try and estimate any L or C of load)
Note there are different inductor models, not sure whether the one in the simulation is the simple version or not (there are more advanced non linear models that deal with saturation)
I wouldn't worry about this too much right now, just get a feel for it all first before complicating matters (just be aware that the simulation inductor may work fine at ridiculous currents etc, try not to do anything too "unrealistic")
Try and use it to "visualise" the theory, understand what is happening and why. Referring to any App notes, datasheet waveforms, etc and comparing might help you keep on track here. Remember it's a slightly different part (not sure exactly what is the difference, the datasheets will tell you though) so keep this in mind too.
If you have your part setup and the necessary tools you could compare simulation waveforms with the real thing.







More... (looser matching)
- Last day of these posts
- In 2011 , 2012 only
- Today
- New search...