Author Topic: Is here way to increase efficiency of wide input boost over 95% ?  (Read 1557 times)

0 Members and 1 Guest are viewing this topic.

Offline MiyukiTopic starter

  • Frequent Contributor
  • **
  • Posts: 908
  • Country: cz
    • Me on youtube
Hi folks,
for small solar project I need a DC/DC to feed common 120V DC bus

I just want to confirm there is no better way of converting 24-100V input to this 120V
Input current is up to 10A (as most panels use this Isc)

I have proposed two phase boost and to reach reasonable efficiency I already have 28 mΩ mosfets and for theoretical 32V 8A input I can get up to 95% efficiency

Is here any way or better topology to solve this wide input regulation, no insulation needed, to get few more percent without increasing price significantly ?
 

Offline Vovk_Z

  • Super Contributor
  • ***
  • Posts: 1478
  • Country: ua
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #1 on: April 24, 2020, 01:38:28 pm »
I'm not sure but 95% is quite a good efficiency as for me.
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 8276
  • Country: de
  • A qualified hobbyist ;)
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #2 on: April 24, 2020, 01:41:56 pm »
You can try to optimize the switching frequency for the inductors (core and wire losses). Maybe it's possible to replace the diodes with some "ideal diode" controller plus MOSFETs.
 

Offline MiyukiTopic starter

  • Frequent Contributor
  • **
  • Posts: 908
  • Country: cz
    • Me on youtube
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #3 on: April 24, 2020, 03:22:16 pm »
You can try to optimize the switching frequency for the inductors (core and wire losses). Maybe it's possible to replace the diodes with some "ideal diode" controller plus MOSFETs.
Diodes have no much room for optimizing as I already have oversize schottky diode with Uf about 0.6, gate drive will consume more than are conductive loss now
Inductors can have some room with adding copper and increase core gap, but it is just a fraction, now is in them about 2x 2.5W (about 1,5-2W core, rest is copper), so I dont think it worth 

Its more like to find other solution if is any or say it is best what can be done.
 

Offline Vovk_Z

  • Super Contributor
  • ***
  • Posts: 1478
  • Country: ua
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #4 on: April 25, 2020, 06:11:34 am »
Quote from: Miyuki
Diodes have no much room for optimizing as I already have oversize schottky diode with Uf about 0.6, gate drive will consume more than are conductive loss now
- if it is so, then there is the same problem with power swich mosfets too?
Do you have your power losses budget (for all parts)?
« Last Edit: April 25, 2020, 06:14:37 am by Vovk_Z »
 

Online jbb

  • Super Contributor
  • ***
  • Posts: 1265
  • Country: nz
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #5 on: April 25, 2020, 06:42:40 am »
Hmm

95% is pretty good. You'll have to work to improve on that.

For hard switching, a synchronous rectifier is, as you say, marginal.  There's not a lot of conduction loss to save, and the SR MOSFET body diode reverse recovery will increase switching losses.

From some basic calcs:
  • 32V * 8A = 256 W in
  • 256 W * (1-95%) = 12.8W loss
  • L Core loss: 2* 1.75W => 27%
  • L Copper loss: 2* 0.75W => 12%
[li]
[/li][/list]

Looks like core loss is a significant factor.  Assuming your material doesn't turn to crap (I'm looking at you, powdered iron), going to a higher switching frequency could allow you to drop the flux density and reduce core loss a bit.

You could have a look at a 3 level boost converter with a flying capacitor.  It needs 4 MOSFETs - which is more trouble - but you can try lower MOSFET V ratings (improves switching losses) and halve the Volt-second product applied to your inductors.

How about some soft switching goodness?
  • Moving to critical conduction mode will reduce turn on losses but might end up increasing core loss(variable switching frequency)
  • Moving to critical conduction mode with synchronous rectification and a little bit of negative inductor current and could deliver Zero Current Switching (variable switching frequency)
  • An auxiliary active soft turn-on stage could get you there.  Aux switch would be Zero Current Switching and the main switch would likely move to Zero Voltage Switching (constant frequency)
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #6 on: April 25, 2020, 08:08:22 am »
Is the input a solar panel? If so, why such a large input voltage range? Is extracting the tiny power at very low voltages, at good efficiency, really important?

Consider adding even more phases and dropping phases at low power.
 

Offline MiyukiTopic starter

  • Frequent Contributor
  • **
  • Posts: 908
  • Country: cz
    • Me on youtube
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #7 on: April 25, 2020, 05:54:14 pm »
Is the input a solar panel? If so, why such a large input voltage range? Is extracting the tiny power at very low voltages, at good efficiency, really important?

Consider adding even more phases and dropping phases at low power.
Input is a solar panel or more in series, I need multiple small DC/DC converters to feed single bus as panels are at different angles and with possible shade
I can live with this efficiency if increasing it will be to complicated

Going to even more smaller phases can be a way just will need to change controller/mcu

Chokes are on ferrite core and switching is at 200kHz as to keep them reasonable small, now at 20uH
Diode conductive losses are about 2x 0.8W
And transistors are about 2W condutive and 1.5W switching

I tried lower frequency 100 or 50 kHz, but chokes get physically huge
 

Offline splin

  • Frequent Contributor
  • **
  • Posts: 999
  • Country: gb
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #8 on: April 27, 2020, 12:27:05 am »
What is your budget for the inductors? I plugged your numbers, (200kHz, 24V to 100V Vin, 120V @ 1.01A out) into Coilcraft's design tool:

https://www.coilcraft.com/apps/power_tools/dc-dc/

Sorted by 'total losses' it offered the 100uH AGP4233-104 with total losses of only 290mW. That loss is calculated at Vmin = 24V so change Vmin to 32V and rerun the calculation. The losses are now a bit higher at 354m for the same inductor. That would save you 2 x (2.5W - .354W) = 4.3W or approx 1.75% which is significant.  Those iinductors are expensive of course at $7.84 and not exactly small: 43.2 x 36.8 x 22mm.

A much cheaper option is the $1.17, 12uH DO5040H-123 with total losses of 371mW which is still a very substantial saving. It's much smaller too (22.35 x 16.25 x 12mm). But, at only 12uH, it would be operating in discontinuous mode which may be a problem EMI/noise wise. They aren't shielded either.

To remain in CCM at all input voltages, the 47uH version, DO5040H-473 could be used but has higher losses - 701mW at 32V Vin, rising to 1218mW at 24V.

How reliable the tool is however is not clear. If you select an inductor (tick the box to the left of the part no,) in the results list and click the green 'Analyze and Graph' heading it shows losses as a function of Vin which don't match the losses shown when varying Vmin in the parameter selection stage.

I believe Wurth and others have similar design tools. If you are looking at for ultra cheap Chinese inductors then getting good information on core losses will probably be difficult to impossible without testing them yourself.

What MOSFET did you select exactly?

PS. If this is a one off project, why not use different inductors for different input voltages? Do they have to be identical?
 

Online trobbins

  • Frequent Contributor
  • **
  • Posts: 826
  • Country: au
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #9 on: April 27, 2020, 01:11:59 am »
Have you made a loss budget and confirmed the budget with measurements?  I'm guessing you would have some parts such as the inductor having a mix of known and uncertain loss contributors, but by subtracting other confident loss contributions from your total loss, then you can bound the uncertain loss levels.  It's always good to start from a known base and benchmark set of conditions, and then swap out one part at a time to identify if a step forward or back has been made.

Is the budget based on working near to the MPPT of your module(s) under close to full sun at their set inclination and seasonal solar elevation and ambient related temp? 

Is the converter acting as an MPPT?
« Last Edit: April 27, 2020, 01:13:56 am by trobbins »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #10 on: April 27, 2020, 02:18:26 am »
At which point does increasing the cost of the converter reduce overall economy?  That is, will the 5W saved by pushing to 97% be worth the 500 cm^2 of panel area saved?

Are there other methods to improve efficiency or economy (better panel cooling? better optimized mechanical mounts, wiring, routing, fusing, busing?), and what is their relationship in the same tradeoff?

Likewise for NRE, is the design time worth it, amortized over product life?

Come to think of it, at current prices, going to 97 or 98% is probably quite justified.  99% or more, you'll be hard pressed to approach, expected design cost going up hyperbolically...

As for techniques, you may find it's worth using a lower frequency, to keep switching and core losses down.  You can afford to use rather large capacitors and inductors here.  Quasi-resonant commutation may be helpful, but don't expect good yields on switching energy (like, maybe you can reduce switching loss by 50%, 80% even, but eventually you will be limited by the high frequency losses of the snubber components, which have to operate at high rates of change).

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline MiyukiTopic starter

  • Frequent Contributor
  • **
  • Posts: 908
  • Country: cz
    • Me on youtube
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #11 on: April 27, 2020, 07:51:00 am »
You are right with inductors, it can be much optimized with those ones.
And also budget wise it have no reason as panel cost is so low it wont make any reason to increase converter price (as panels itself is now about 0.3-0.4 $/Wp)

I think Ill go to 2 phase at 250kHz with
BSC500N20 transistor should have about 1.3W cond + 0.7W switch losses and is cheap
Cheap Ferrocore inductor two in series 15uH; 6A; 22mΩ - simmilar to Coilcraft MSS but cost just 0.4 $

All controlled with SAMD11

It seems to me as reasonable compromise between efficiency, BOM cost and time to put together
« Last Edit: April 27, 2020, 07:55:22 am by Miyuki »
 

Online trobbins

  • Frequent Contributor
  • **
  • Posts: 826
  • Country: au
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #12 on: April 28, 2020, 11:57:27 am »
Will you be posting your breadboard results?  It would be good to see how close you got with your efficiency estimate, especially as you are operating at 250kHz.
 

Offline MiyukiTopic starter

  • Frequent Contributor
  • **
  • Posts: 908
  • Country: cz
    • Me on youtube
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #13 on: April 28, 2020, 12:09:33 pm »
I am also curious how it will behave and how many times it will just release magic smoke violently to my face  :-/O
Hope Ill have a layout during this week
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #14 on: April 28, 2020, 01:17:13 pm »
Do these "Cheap Ferrocore inductors" have core loss models / web software available?

AC loss can sometimes surprise you, even with ferrite-cored inductors. If you make an assumption of it being zero (or some percentage of DCR loss), you may succeed by luck, or not. Selection software like Wurth's Redexpert I like to use compare the total losses of parts, and you can sometimes see a certain part has 90% DC loss, 10% AC loss while some other part has the opposite! Total is, of course, the only thing that matters.

This being said, cheap, widely available inductors with no AC loss specifications are often worth of testing in prototyping. If you run them at rated DC current (with specified 40degC temp rise, for example), and notice that the actual temperature isn't that much more (i.e., measure 50 degC temp rise in your switcher), you can quickly see you were lucky regarding the AC loss.

I like to use "generic" SMD inductor footprints which have large pads so I can fit many sizes and types of inductors. And pick a footprint one size up from the inductor I think I will be using.
« Last Edit: April 28, 2020, 01:19:44 pm by Siwastaja »
 

Offline MiyukiTopic starter

  • Frequent Contributor
  • **
  • Posts: 908
  • Country: cz
    • Me on youtube
Re: Is here way to increase efficiency of wide input boost over 95% ?
« Reply #15 on: April 28, 2020, 01:31:24 pm »
Do these "Cheap Ferrocore inductors" have core loss models / web software available?

I expect them to be in same order of magnitude as Coilcraft ones as they have same sizes and similar Rdc and Isat
So Ill use Coilcraft calculator, find closest cheap one and say double AC losses to get worst expected case
If coil desolder from board Ill see they are much worse, but there is no much to invent here
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf