LM5145 Inductor keeps heating up

[MBrahim]

I designed a 42V -> 12V buck converter that can drive up to 20 current.
when I got the PCB and assembled it, the inductor keeps heating up at no load and I don't know why.
I looked for it and people said check for core loss but i can't find it anywhere in the datasheet
https://www.lcsc.com/datasheet/lcsc_datasheet_2304140030_Chilisin-Elec-MHCC12050-1R0M-R7_C329553.pdf
I used MHCC12050-1R0M-R7
it's maximum ripple current said to be 40%, i designed it to be 44% but i don't think that is the problem, the heat builds up like crazy, I can't even debug it with OSC as I keep disconnecting it before it burns.
I looked again to datasheet and found out that maximum voltage is 30VDC, I never saw this one before on any coil i have ever used :'(
Is there a way to overcome this problem?
If I added a zener clamp to the input will it work?
or what should i do?

[Someone]

A 20A converter on 22uF output caps, with a 12mm inductor, slow low rent mosfets, gunning for a 1MHz switching frequency, what were they thinking?

Everything about it looks suspect. Buy the evaluation board and start from a working design.

[mag_therm]

Rt = 10 k  F=1e6 Hz ?
L1 = 1 uH ?

Then XL1 = 2*pi*F*L1
=j6.28 Ohm

Input = 42Vdc ?
Eacin=.....
IacL1=...

[T3sl4co1l]

What were your core loss calculations?

Trick question, I know, the datasheet doesn't give core loss data at all.

That's a warning sign.

40% is a lot of ripple for an inductor without core loss specs, and is still a lot even for a gapped ferrite type.

Composite/molded types can perform quite well, but the material comes in a range of losses and it's not clear which kind they used here.  Perhaps it's the cheaper / older-generation higher-loss material?

A much larger, lower-loss inductor is most likely to help here.  At least with the immediate inductor-heating problem.

Perhaps there are layout or component-selection issues, or the transistors aren't well suited to the driver, but we don't have information to assess that as yet.

A 20A converter on 22uF output caps, with a 12mm inductor, slow low rent mosfets, gunning for a 1MHz switching frequency, what were they thinking?

What's wrong with the transistors?  They look fine: https://wmsc.lcsc.com/wmsc/upload/file/pdf/v2/lcsc/2105241820_ALLPOWER-ShenZhen-Quan-Li-Semiconductor-AP68N06G_C2828590.pdf
I think a lot of people are stuck in the past and haven't realized how good transistors are coming from Taiwanese brands, and even the mainland.  You still might have the issue of sourcing legitimate parts, but that's true of any brand purchased in the area, and there are channels to go through, and means to determine such.

For my part, I've tested some no-name (from a western perspective) SuperJunction parts that perform equally well to brand name (e.g. Infineon) parts.

Tim

[mag_therm]

Core loss calcs won't help that L1 if it is indeed 1 uH
That means AMPs ( not mA) of Ac ripple at 1 MHz

I have a similar Dc Converter : 36V solar to 12V 20Amp
My pwm 7 time lower than MB's  140 kHz;  I used 7700 uH    Xl = j6700 Ohm
Ac ripple voltage in to L1 =0.7 * 36  = 25 V approx
Ripple current in L1 = 25/6700 = 3.4 mA  ( It is a battery charger)

[strawberry]

according to datasheet 20A inductor must be at least 40C @25C
2.5mOhm * 20A^2 = 1W

[T3sl4co1l]

Core loss calcs won't help that L1 if it is indeed 1 uH
That means AMPs ( not mA) of Ac ripple at 1 MHz

Well yeah they said 44% calculated.  44% of 20A is 8.8Apk.

I have a similar Dc Converter : 36V solar to 12V 20Amp
My pwm 7 time lower than MB's  140 kHz;  I used 7700 uH    Xl = j6700 Ohm
Ac ripple voltage in to L1 =0.7 * 36  = 25 V approx
Ripple current in L1 = 25/6700 = 3.4 mA  ( It is a battery charger)

Is that powder? That's big even for laminated iron.  What control scheme did you choose?

Tim

[MBrahim]

A 20A converter on 22uF output caps, with a 12mm inductor, slow low rent mosfets, gunning for a 1MHz switching frequency, what were they thinking?

Everything about it looks suspect. Buy the evaluation board and start from a working design.

First of all if you don't mind don't just move and trash talk anyone, We're not born with soldering Iron in our hands sir.
But I'll answer you anyway.
I was thinking according to TI WEBBENCH designer and the buck converter calculator that They provide in thier site

About the components selection, Coil was a mistake and I admitted it as I got no help from the education system on how to choose your coil or what are the parameters you need to check, It was all self-learning with limited resources. The Output caps are chosen according to the designer as well, every little detail even the layout I did it according to the datasheet with a little edit for my project needs.

And for that part about buying the evaluation board like bro in what world are you living, it's expensive, shipping can take a while. I needed to design very fast, there're deadlines.

[trtr6842]

Inductor datasheets are notoriously sparse on core loss information.  Coilcraft has one of the better tools.  I plugged in your converter parameters into their core loss comparison tool with the XAL1060-122.  Its a similarly sized inductor and it's showing 3W of core + AC winding loss at 42Vin, 12Vout and at 1MHz switching frequency.  I suspect the losses in your inductor are at least that or even higher given the fact that it's saturation current is even lower than the coilcraft part.

Try running your converter at 15Vin and 12V out.  That will cut the core loss way down (8% of whatever you're getting at 42Vin).  This should buy you some time to actually scope out the "SW" node, output voltage, etc.  If you can get that scenario to work, then you can start bumping up the input voltage and you'll see what happens.  Keep records of input and output voltage and current and you'll end up with a good idea of how the overall converter losses change based on those parameters.

I think you're way more likely to have success if you reduce your switching frequency to 150-300kHz.  You mentioned TI's Webbench designer, if you look at all their suggested designs for 42Vin and 12Vout they are mostly in that range.  Core and MOSFET switching losses will be way more manageable.  You will need to increase the inductance, get the 4.7µH and 10µH versions of that inductor and give them a try.  You won't be able to get 20A out of them, but it should let you at least get this design to do something and will help you learn as much as possible for the next revision. 

I'm afraid there's no magic bullet that will save this design as is.  You might be able to find a much taller inductor that will get you closer to the design goals, but my guess is that part will need to grow significantly.

You can also look into winding your own toroidal core inductors, that could be good for prototyping and figuring out what is required to reach your design goals and select an inductor for production/end use.

[mag_therm]

Is that powder? That's big even for laminated iron.  What control scheme did you choose?

Tim

My core is a surplus 3C85 double C . Outside dimension about 4 inch by 4 inch,
I made two core tubes of brazed 6160 with a gap and grounded. Two windings each of 2 by 18 G in parallel, connected in series to give 44T.
The temp rise was within Class B on 20 A test.
The DC converter has no base frequency filter caps, the battery equivalent C gives a double pole at less than 1 Hz I think.
There is a pi section RF filter with a small lossy iron core at output.
One reason the inductance is so high, is that the 13.8V load ( ham radio) is often less than 3 Amp,
If the load current in the choke becomes discontinuous the regulator is not stable.
Volt Error amp is type 3

My output filter with maybe -50 to -60 dB attenuation in proximity to HF radio is very different to MB's present design.

I see that the LM 5145 data sheet calls for L1 giving  30% ripple. But at 1 MHz I wonder if practical.
I think the track inductance + ESL of filters (total 100nF)  might be 10% of the L1
with resulting high ripple at 1MHz base freq.
And the inductor turns will be heated by that 1MHz ripple of 6 Amp; reference depth is 72 micrometre giving poor copper utilization.
Maybe Litz would be needed. Outside of my experience, but I think the eventual inductor will be physically much larger than his photo shows,
and possibly the 30% ripple current recomedation will have to be reduced, and maye NiZn ferrite used.

[mag_therm]

I found some photos of building and testing the filter choke in the 36 ~ 42 V in : 13.8 V 20 A converter back in 2022
Winding:
https://app.box.com/s/h46uat4d1nc08ltr29rxegas7rhogcro
Complete Inductor:
https://app.box.com/s/a74rhrcfat9hfbrsisjyppdyjw7okcg4

Load Test with Spectrum Analyser tap:
https://app.box.com/s/zzr2vde1wnqqwet9jyttt45znl7qan3d

1 ~ 80 MHz spectrum:
https://app.box.com/s/dmjn65bqwrran244ns07gerbjgbnffa7

Let us hope MBrahim ends up with a somewhat smaller (!) inductor with similar performance.

[T3sl4co1l]

Try running your converter at 15Vin and 12V out.  That will cut the core loss way down (8% of whatever you're getting at 42Vin).  This should buy you some time to actually scope out the "SW" node, output voltage, etc.  If you can get that scenario to work, then you can start bumping up the input voltage and you'll see what happens.  Keep records of input and output voltage and current and you'll end up with a good idea of how the overall converter losses change based on those parameters.

Good idea.  Note that the XAL parts are probably low loss as powder go, and the part in question here may be still higher loss.  Combined with the small size, that pretty well explains the rapid temperature rise.

I think you're way more likely to have success if you reduce your switching frequency to 150-300kHz.  You mentioned TI's Webbench designer, if you look at all their suggested designs for 42Vin and 12Vout they are mostly in that range.  Core and MOSFET switching losses will be way more manageable.  You will need to increase the inductance, get the 4.7µH and 10µH versions of that inductor and give them a try.  You won't be able to get 20A out of them, but it should let you at least get this design to do something and will help you learn as much as possible for the next revision.

Higher inductance seems to be the top priority, and lower frequency can help as well.  Of course the converter will get much larger in overall footprint, but if the pressure is to have something that works at all, this seems like a useful direction, and then it can be optimized further as time and budget allows.

The controller I guess is voltage mode (yuuuck!), and the application section pretty clearly recommends "30-40%" current ripple, but declines to explain why.  If it were a peak-current-mode controller with slope compensation factor of 2 to 3, this would be a lower stable limit, but voltage-mode can simply be made arbitrarily low and it's fine, given the compensation is adjusted accordingly.

Even without adjusting Fsw, increasing L and adjusting compensation should help greatly.

A moral of the story: design tools are just that, tools to get started, or to get somewhere or do something, given that you know what you're doing.  Nothing can replace an engineer. Not yet anyway; maybe in a couple decades, but not right now.  Tools will gladly give you ludicrous or nonsensical results, or results that are just a starting point based purely on limited parameters such as component values and ratings, and then after that, it's up to you to select components that can actually do that without melting themselves off the board, or to meet space or cost or other limitations.

I'm afraid there's no magic bullet that will save this design as is.  You might be able to find a much taller inductor that will get you closer to the design goals, but my guess is that part will need to grow significantly.

You can also look into winding your own toroidal core inductors, that could be good for prototyping and figuring out what is required to reach your design goals and select an inductor for production/end use.

Nice thing about low ripple fraction, powdered iron toroids become practical, even quite high loss materials like #52.  Though I wouldn't choose that material for Fsw over 100kHz, really, but other materials will do just fine.

Litz I think becomes useful somewhere north of 40% ripple; it depends what balance of losses you get (or need) in the thing, but it's definitely of use when high efficiency is required, and BCM or DCM is used.  Around there or below, a few strands, or round wire of just somewhat larger than calculated DCR size, is adequate for the job.  You never see litz on powdered iron of course, for obvious reasons.*

*Well, commercially speaking, lol.  I did once make this,
https://www.seventransistorlabs.com/Images/Magamp_PSU2.jpg
a boost converter using a #2 powdered iron core (low mu, low loss) I had on hand, plus ~10AWG equiv. litz, which I also had on hand.  Well, even running in DCM, it simply doesn't get hot.  It's also running all of like 5A DC.  Ridiculously overbuilt, as spare-parts solutions sometimes are, heh.

Tim

[MBrahim]

well i got a question, all of the comments ignored tha part that the inductor limit is saying that the maximum VDC is 30V, it's the first time i have ever seen a voltage limit(atleast that low) in an inductor datasheet.

another thing, I tried to replace it with 3.3uH inductor and the surprise is that it heated but slower, but still after time more time (not that much) I can't even touch it like the first inductor did. so what can i expect from that?

[trtr6842]

The voltage rating is a dielectric rating.  No inductor has some natural voltage limit, but depending on the construction there will be a limit to the insulation between coils, or from coil to core.  These molded ones are designed for lower voltage applications, so the internal coils must not have great electrical insulation.  What that means for your design is that there is a chance the coils could be shorting to each other, or to the core. 

It's not surprising that the 3.3uH inductor has less loss.  That inductor, if it's make of a similar material and same size, must have more turns on it's internal coil than the 1uH one.  That reduces the flux density swing in the core, which means lower core loss.

Again, I highly recommend trying to run your converter with a lower input voltage.  This should let it actually run and take scope measurements.  There could be other hidden issues, and it would be helpful to rule those out early.

[Someone]

A 20A converter on 22uF output caps, with a 12mm inductor, slow low rent mosfets, gunning for a 1MHz switching frequency, what were they thinking?

Everything about it looks suspect. Buy the evaluation board and start from a working design.

First of all if you don't mind don't just move and trash talk anyone, We're not born with soldering Iron in our hands sir.
But I'll answer you anyway.
I was thinking according to TI WEBBENCH designer and the buck converter calculator that They provide in thier site

About the components selection, Coil was a mistake and I admitted it as I got no help from the education system on how to choose your coil or what are the parameters you need to check, It was all self-learning with limited resources. The Output caps are chosen according to the designer as well, every little detail even the layout I did it according to the datasheet with a little edit for my project needs.

And for that part about buying the evaluation board like bro in what world are you living, it's expensive, shipping can take a while. I needed to design very fast, there're deadlines.

Did WEBBENCH provide the mosfet part numbers? or the inductor? those appear to be your (poor/misguided/uniformed) choices. As you've not posted the produced design I have no idea if those capacitors were what was recommended either (seems unlikely). You're out of your depth from what you've described.

Not buying the dev board will probably cost you more in time and revisions to get this up and running. You could meet your deadline in all sorts of ways, but demanding the public do it for you for free? lols

[Someone]

WEBENCH design for those interested:

[tooki]

Remember that in WEBENCH, you can click on a component (either in the schematic or in the BOM, I forget which) and see not only what part it selected, but what filter criteria it used to make that selection.