How to select inductors for a dual conversion circuit?

[kraal]

Hello,

I'm experimenting with LT8330 inverting converter, and have some questions about inductors selection/placement.

Here is the context:
The datasheet [1] provides examples of typical applications of the chip in a boost/SEPIC/inverting converter configuration.
These examples (SEPIC/inverting) involve the usage of two inductors. When looking at the specs of recommended inductors, all of them are "coupled inductors" (with 2 windings).
None of these examples show how to achieve dual (+/-xV) conversion, but I found a technical article from Analog Devices [2] providing an example of +/-15V conversion.
Here is the corresponding diagram (figure 8 of the article):



This example makes use of 3 inductors' windings and recommends the usage of an Eaton's VP4-0075 [3] (6 windings).

The datasheet says nothing specific about any kind of inductors "matching".

My questions are the following:

1) do I have to use such an inductor (the VP4-0075 has six windings, and i only need three for this circuit. This looks like a waste to me)
2) can I use a two-windings inductor (L1A + L1B) + a single-winded inductor (L2) for such a circuit ?
3) if 2) is possible (with no negative impact), on a diagram such as the one above, does it matter where I put "L2" (the one with a single winding) (i.e. L1A, L1B or L1C position) ? (and if yes, what would be the best configuration)
4) What would be the impact of using two two-windings inductors (L1A + L1B and L2A + L2B) for this circuit ? (which would mean to not use a single winding of one of the two inductors)

Thanks a lot for your help !

[1] https://www.mouser.ch/datasheet/2/609/LT8330-1660250.pdf
[2] https://www.analog.com/en/technical-articles/low-iq-60v-monolithic-boost-sepic-inverting-converter.html
[3] https://www.mouser.ch/datasheet/2/87/eaton-vp-versa-pac-inductors-transformers-data-she-1608793.pdf

Edit: fixed some typos

[duck_arse]

it's strange they use a six winding thing and then make no mention of whether or not they use the extra three windings. the datasheet does provide some hints.

I guess you could wind three coils on a form/ferrite and give it a smoke test, but they all need to be on the one core. easiest way is to get three strands of ECU, mark one end of each strand, and wind them all at the same time, trifilar. then use the marked end as the dot connection/s.

[kraal]

Thank you for your answer.

the datasheet does provide some hints.

Indeed, the datasheet states "As shown in Figure 5, the SEPIC converter contains two inductors: L1 and L2. L1 and L2 can be independent, but can also be wound on the same core, since identical voltages are applied to L1 and L2 throughout the switching cycle."

Apparently they do not want to take a stand, nor provide constraints involved by each option.
Should I understand it as "do as you wish, there will be no impact in this configuration" ?

I guess you could wind three coils on a form/ferrite and give it a smoke test, but they all need to be on the one core. easiest way is to get three strands of ECU, mark one end of each strand, and wind them all at the same time, trifilar. then use the marked end as the dot connection/s.

I think you probably have more confidence in my skills than I have 

[Phoenix]

What's shown there is a "coupled inductor Ćuk/SEPIC converter" (two different types of buck/boost converter, one inverting, the other not), which uses transformer action of the coupled windings to reduce output ripple. You can use independent inductors at the expense of higher output voltage ripple (all else being equal).

However, do you really need all the features this type of circuit offers? It's a buck-boost converter, meaning it can produce voltages either greater or lesser in magnitude than the input (as well as being able to invert voltages for negative rails). For most pedal-type applications, or even just DC-DC applications in general, the input voltage will always be either greater or lesser than the desired output, so either a buck (for the former) or boost (for the latter) is normally all that is required, which simplifies the design, and is also more efficient. One would only normally use a buck-boost converter if a product had to work over a wide range of input voltages (not the normal case for pedals).

If I were starting off trying out some DC-DC circuits, buck-boost is not where I would start personally.

Edited - errors.

[Phoenix]

Oh, and the specified 6 coil inductor uses all 6 windings: 3 pairs in parallel.
12.7µH||12.7µH ≈ 6µH as listed on schematic.

[antonis]

Should an e-mail on Analog Devices technical support be more enlightening..??

[kraal]

Hello and thank you for your reply.

However, do you really need all the features this type of circuit offers?

To answer your question, I have no specific need for now, I just want to experiment, understand, with the objective of being able to add one more tool to my "toolbox" and, in the future, know when to use it (if needed).

Oh, and the specified 6 coil inductor uses all 6 windings: 3 pairs in parallel.
12.7µH||12.7µH ≈ 6µH as listed on schematic.

The V04-0075 used on the schematic is 6uH, and I thought that the symbol they use on Figure 8 is for an Iron core (with polarity). 2x 6uH in // would do 3uH isn't it ? Am I missing something ?

Regards

[kraal]

Should an e-mail on Analog Devices technical support be more enlightening..??

Probably more enlightning than your reply 

More seriously: Phoenix's answer was very instructive, it forced me to go read more thoroughly, think about what I want to achieve, and confirmed why I'm trying to do this. It's also part of the learning process / path don't you think ? I don't think that a technical support would provide this.

Regards

[Phoenix]

The V04-0075 used on the schematic is 6uH, and I thought that the symbol they use on Figure 8 is for an Iron core (with polarity). 2x 6uH in // would do 3uH isn't it ?

Am I missing something ?

Only that I shouldn't post when half asleep.

I was looking at the wrong part on the datasheet (VPH4, not VP4).

But... I was thinking series/parallel inductors on seperate cores, so my math was wrong anyway. From the datasheet:
"Paralleled VERSA-PAC windings result in no change to the net inductance because the total number of turns remains unchanged; only the effective wire size becomes larger. Two parallel windings result in approximately twice the current carrying capability of a single winding."

As for iron vs ferrite core, representing this in the schematic symbol is less strictly adhered to. In fact, notice that in figures 4, 5 and 6, they're all drawn as "iron core", yet must all be ferrite because of operating frequency, and is confirmed through listed part numbers. Notice though that for figures 4 and 5, only a single bar is used in the inductor symbols, and the specified parts are dual winding, but in figure 6 each inductor symbol has two bars, representing 6 windings total.
This is a terrible way to represent parallel windings, as the bar is supposed to represent the core (and ideally give some indication of core type), not "number of windings", but magnetics have some of the most abused schematic symbols because they're less commonly used, and they're the least ideal part, so less practical information can be communicated through a symbol than for a resistor or cap. A part number is really required in most cases, which conveys all the real info. So long as the schematic conveys the phase information correctly that's all that really matters, part number conveys the rest.

[Rob Strand]

When you parallel equal windings on the same core it doesn't change the inductance.   It makes the DC resistance lower.  It should also reduce the AC resistance.   (The transformer datasheet has some hints of this on page 8.)

[duck_arse]

Thank you for your answer.

the datasheet does provide some hints.

Indeed, the datasheet states "As shown in Figure 5, the SEPIC converter contains two inductors: L1 and L2. L1 and L2 can be independent, but can also be wound on the same core, since identical voltages are applied to L1 and L2 throughout the switching cycle."
......

I actually meant the datasheet for the transformers, right at the end has windings/connections info. my eyes had glazed over by that stage, tho. listen to those other guys. [not antonis, obviously.]