Lowest profile boost 5V, 2A to 18V, 0.5A

[levnu]

where can i find the lowest possilble such boost?
what will be its total thickness?

if the inductor is the highest component can i replace it with few low profile inductors in serial?
if yes what will be the lowest profile boost that i can achieve?

the efficiency of the boost is not importent for me!

 

[emontllo]

Hello,
Do you want a complete boost module or you want to design it?
As you well said, inductor will typically be the highest component. Capacitors can be easily paralleled and are available in low profile.

To achieve minimum height I would suggest building planar inductors embedded on PCB. The inductor coils will be PCB traces:
Some info:
https://en.wikipedia.org/wiki/File:Planar_core_assembly_exploded.png
https://www.youtube.com/watch?v=qPpYDlLpjyk

Good luck!

 

[levnu]

thanks for fast response
i want to use the next boost but somehow to find lowest possible inductor
https://www.ti.com/product/tps61175
Do you think i can find 10uH lower than 2mm thicknes that will stand with at least 2A?

yet your idea for planner inductor looks intersting

 

[emontllo]

Have a look to: Vishay Dale: IHLP2020BZER100M01
**broken link removed**

2.3 A rating, 4 A saturation, 2 mm height, available in Digikey. This one can work.

Also consider putting 2 lower inductance inductors in series... this may not be the best for a high frequency boost but it could save some height.

When building the PCB stackup also consider building a thin PCB, it is easy to find 0.8 mm thick FR-4 PCB with 4 layers.

 

[levnu]

Thank you i found some inductor that seems i could use 2 in serial to achieve 10uH, but why you say this may not be the best for high freq boost?
https://katalog.we-online.de/pbs/datasheet/7440700047.pdf

can i use it on 2 layers 0.4mm FR-4 PCB?

- - - Updated - - -

Thank you i found some inductor that seems i could use 2 in serial to achieve 10uH, but why you say this may not be the best for high freq boost?
http://katalog.we-online.de/pbs/datasheet/7440700047.pdf

can i use it on 2 layers 0.4mm FR-4 PCB?

 

[mtwieg]

Putting inductors in series will cause the node in between them to oscillate significantly (unless you damp it appropriately), which can be a problem from an EMI perspective. Putting inductors in parallel doesn't have this problem, though. PCB trace inductors will radiate a great deal of EMI, and take up a lot of board space.

 

[levnu]

Thanks for now i do not think of the EMI yet maybe i can use only one 4.7uH inductor?
i'm sorry for asking so much question below, it just cause i don't get it yet.

1. in the application note i saw i may use a 4.7uH inductor (1.8mm thickness!!! & around 50mOhm DCR) it says that its the lower limiit...
so can i use only the inductor in the link: https://katalog.we-online.de/pbs/datasheet/7440700047.pdf ???
2. my Vin is 5V DC, 2A DC and out will be 18V, 0.5A, yet i wonder how i deal with the Delta ILoad? how i deal with the current ripples?
becasue in this specific inductor the saturation current is 2.2A and i might have higher ripples around the 2A input current
3. another issue that i didn't get is if i need to work on DCM or CCM state?
4. and how can i set the duty cycle?
5. i know the R1 & R2 that i need for achiving the 18V output while my input is 5V, but will it be ok with the 4.7uH inductor or should i change that Duty cycle for it to be ok,
and if yes so how? does changing the Duty cycle can even help me to deal with the 4.7uH inductor? instead of using the 10uH which is very high 4.5mm thickness
6. how i determine the 0.5A output load?

many thanks and sorry again for so many questions

 

[mtwieg]

Most of your questions will depend on the controller used for the supply. Operating at higher frequencies will allow for smaller inductors and capacitors. You should probably try to operate at 200KHz or more, so that you can maintain CCM with small inductance. For example if you operate at 1MHz then a 3.3uH inductor will give ~1.1A of ripple current. The controller IC should take care of duty cycle control assuming you use it correctly. For output capacitors, you will likely want to use ceramic capacitors, probably in packages like 1206, 1210, 1812, etc in low profiles.

 

[levnu]

why increasing the Fswitch helps me to reduce inductence?
i would like to work with the step up below:
https://www.ti.com/lit/ds/symlink/tps61175.pdf 240KHz its minimum freq...
will the Duty cycle will be determind by its internal PWM? it is not somehting i choose?
will it be ok with the 4.7uH inductor in my post above? what mechanisem makes it be on CCM or DCM?
what cause the output current ripples my low inductance?
thanks

 

[mtwieg]

Increasing switching frequency will reduce ripple current in the inductor. Therefore if you increase frequency, you can decrease inductance in order to maintain the same ripple current.

That controller should work fine. The duty cycle will be automatically adjusted by the feedback loop to maintain output regulation. Though you can roughly calculate what it will be at steady state, given the various parameters of the converter.

 

[BradtheRad]

why increasing the Fswitch helps me to reduce inductence?
i would like to work with the step up below:
https://www.ti.com/lit/ds/symlink/tps61175.pdf 240KHz its minimum freq...
will the Duty cycle will be determind by its internal PWM? it is not somehting i choose?
will it be ok with the 4.7uH inductor in my post above? what mechanisem makes it be on CCM or DCM?
what cause the output current ripples my low inductance?
thanks

Your plan is possible, for the most part.

Screenshot of simulation:



Notice that your switching device will need to carry upwards of 4 amps. However the TPS61175 has a 3A rating.

-----------------------

To assist in understanding the interaction of coil and frequency and duty cycle, click the link below. It will open the Falstad.com website, load my schematic above, and run the simulation on your computer. (Click Allow to load the Java applet.)



Click the switch at right to begin the power cycle. Let up to see the output cycle.

The idea is to create a pattern of current flowing in the coil so that it will maintain your desired volt level at the load.

You can alter values by right-clicking a component, and select Edit.

I included small resistances to represent the fact that there is some internal resistance in components.

 

[emontllo]

Hello Levnu,

If minimum size is a priority and EMI's are not a concern, I would suggest operating > 1 MHz, try 2 MHz. In this frequency, special care is needed when designing the PCB layout (see datasheet recommendations).
As a rule of thumb: inductor saturation current < (Average input current + current ripple / 2)

Actually, there is a pretty nice WEBENCH to design the converter in the right side of the TI website https://www.ti.com/product/tps61175. This will answer most of your questions regarding specific design components. Set the optimization to: smallest footprint and chose the maximum frequency. Find attached the results it gave to me with 2 MHz --> 8.2 uH inductor.

 

[levnu]

Thank you both for the much help