Author Topic: Limiting initial charge current into supercap from a boost converter?  (Read 1532 times)

0 Members and 1 Guest are viewing this topic.

Offline wb0gazTopic starter

  • Regular Contributor
  • *
  • Posts: 190
(edited to clarify operating use case)

I need to charge a supercap (1.6F, 16V) from a small boost converter (can deliver about 500 mA, and does not have onboard current limiting).

I am looking for simple(st) way to avoid overloading the boost converter while the supercap is charging.

The use case is this - the boost converter is sufficient to supply all of the energy the application requires over it's life cycle, however, there are occasional peak current demands (see clarification below) from the application which I intend to satisfy with the supercap, which would then be recharged from the boost converter. Voltage stability is not required (that is, the supercap's voltage would drop after discharge and would be (relatively slowly) replenished from the boost converter.

(CLARIFICATION - the peak current period is very brief and will reduce voltage across the super capacitor only modestly; this problem is focused on managing initial charge-up of the super capacitor after power-up where the super capacitor would pose a short- or near-short circuit load to the preceding boost converter - so losing some energy - perhaps from heat dissipated in a pass element - is OK as this energy loss would be a one-time event after power-up; after initial power-up the voltage across the super capacitor will be high enough that it won't short-circuit the preceding boost converter)

The boost converter is based on ME2149 IC.

The supply power to the boost converter is not specifically limited, however, I believe that limiting current prior to the boost converter may not be successful as the boost converter (which works in a 2-6V input range and will nominally be supplied with 4-5VDC) does not operate fully with supply voltages below 2V (hence, while the supercap is charging, the supply and boost converter will be working into a very low ESR, particularly at initial start-up.)

Suggestions?

Can I clarify the requirements or circumstances or use case further?

Thank you!

Dave
« Last Edit: June 21, 2021, 01:38:26 am by wb0gaz »
 

Offline sandalcandal

  • Supporter
  • ****
  • Posts: 641
  • Country: au
  • MOAR POWA!
Would be helpful if you can post a schematic of your current (planned) implementation.

Edit: This response is incorrect. See below

Normally in this situation you'd be doing some sort of soft-start with the boost converter running at low duty cycle to limit current, possibly with some current limiting feedback control. Not sure from the datasheet but it looks like it might be possible to do an extended soft start using a capacitor on the CE pin slowly charged up by a resistor to Vcc (might also need a parallel resistor to discharge this capacitor). Bypassing R1 in the feedback divider with a parallel capacitor could also have a similar effect.

You could do a more complicated network with actives on the feedback pin to do more controlled current limiting but then it is much less cheap or simple.

Frequency is fixed and it looks like it has an internally limited maximum duty cycle so you might be able to just limit current to safe levels by picking an appropriate inductance.
« Last Edit: June 21, 2021, 04:17:33 am by sandalcandal »
Disclosure: Involved in electric vehicle and energy storage system technologies
 

Offline sandalcandal

  • Supporter
  • ****
  • Posts: 641
  • Country: au
  • MOAR POWA!
Worst case current (from the basic inductor equation) would be
\[ I_{max} = \frac{V_{max} D_{max}}{L f_{min}} \]
Subbing in max recommended inductance and other values from this datasheet with your given max supply voltage:
\[ I_{max} = \frac{6*0.78}{4.7*10^{-6}*0.8*10^6}= 1.24 A\]
Which is within the datasheet current rating for all the non SOT23-5 versions it seems? Also not clear to me from the datasheet  how the PFM works which might result in \$f_{min}\$ going lower and raising current.

Edit: corrected values and improved wording.
« Last Edit: June 21, 2021, 02:08:46 am by sandalcandal »
Disclosure: Involved in electric vehicle and energy storage system technologies
 

Offline wb0gazTopic starter

  • Regular Contributor
  • *
  • Posts: 190
Thank you sandalcandal for the early reply.

I'm not able to post a schematic right now, however, the entire end-to-end solution looks just like the ME2149 data sheet typical application circuit, except that Rl is in parallel with a 1.6F 16V capacitor with low (and unspecified) ESR.

The boost converter is exactly represented by the datasheet page 2 ("typical application circuit"). I do not know right now the parameters of the inductor L (I'll be working with a small module sourced online), however, it seems that there would be risk to the inductor (at least) when the converter is working into a short circuit (which it would see - at least approximately - at cold start.) From what I've been able to gather, this type of circuit is not intended to work into a direct short (i.e., supplying some finite current into a short circuit.) I am uncertain of this scenario (hence my posting.)

Assuming that the boost converter can't work into a short circuit (posing some in-range current to the supply on the left, and delivering current into an approximately zero-ohm load on the right), I  was wondering if I could identify some circuit that would pose a low voltage drop (unspecified) when current flow is below some threshold (500 mA in my example) and increases when current flow tries to exceed the threshold (so that the total current flow does not exceed 500 mA), that seems like it would solve the problem. The element would dissipate the energy it absorbs during the start-up phase, which is OK.

Thanks again,

Dave
 

Offline wb0gazTopic starter

  • Regular Contributor
  • *
  • Posts: 190
One sentence in the datasheet caught my eye (from page 8, section 3, part of "External parts selection for DC/DC converter"):

"However, a higher capacitance is recommended if the output voltage is high or the load current is large."

I don't know if this means "higher capacitance" would include 1.6 Farads? Even a few tens of uF would be a high load for a tiny period of time, however, this capacitance would be ~100K times that envisioned in the datasheet.

In this application, output voltage will be 12-14 volts and load current of the application net of the super capacitor (although extremely low duty cycle) could be several amperes; load current into the super capacitor would ideally be limited to a safe level for the boost converter (500 mA in this example.)

« Last Edit: June 21, 2021, 02:23:55 am by wb0gaz »
 

Offline Someone

  • Super Contributor
  • ***
  • Posts: 4493
  • Country: au
    • send complaints here
I need to charge a supercap (1.6F, 16V) from a small boost converter (can deliver about 500 mA, and does not have onboard current limiting).

I am looking for simple(st) way to avoid overloading the boost converter while the supercap is charging.
Use a switching regulator with built in current limiting, any external solution will likely end up more expensive than picking the right part to begin with.
 

Offline wb0gazTopic starter

  • Regular Contributor
  • *
  • Posts: 190
Choosing a different converter/regulator is certainly a valid option. This (switching DC energy conversion) is not my primary field, so I am also open to suggestion for device or approach (supply voltage 4-5 volts; delivered power would be in 12-14 volt range limited to 0.5 amperes, tolerant of dead-short load - about 6-7 watts maximum total delivered power to load.)

Also, in doing some further reading, this problem appears to be an "inrush" type problem (probably encountered widely in power supplies driving applications with significant shunt capacitance.)
« Last Edit: June 21, 2021, 03:04:14 am by wb0gaz »
 

Offline wb0gazTopic starter

  • Regular Contributor
  • *
  • Posts: 190
This (from texas instruments) seems relevant:

Application Report
SNVA731–June 2015
Working with Boost Converters

https://www.ti.com/lit/an/snva731/snva731.pdf

It covers basic issues with shorted output and inrush conditions, although the paper does not directly discuss driving a highly capacitive load such as I envision.
 

Offline sandalcandal

  • Supporter
  • ****
  • Posts: 641
  • Country: au
  • MOAR POWA!
Apologies for the poor initial answer. I had my head thinking of other switch mode topologies.

At initial V_cap < V_in, uncontrolled current can indeed simply pass through the inductor and diode, completely avoiding the controller/switch. In order to control current, we require an additional component in series with the inductor or the diode.

A buck-boost or flyback topology may be a better choice for this application.

Other typical inrush control means such as an NTC resistor are probably unsuitable for your intermittent power application.

There are also ICs specifically designed for "capacitor charging" so I would suggest looking into those if you haven't already.
« Last Edit: June 21, 2021, 04:19:48 am by sandalcandal »
Disclosure: Involved in electric vehicle and energy storage system technologies
 

Offline f4eru

  • Super Contributor
  • ***
  • Posts: 1081
  • Country: 00
    • Chargehanger
Buck-boost may be useful.
I used a linear current sink on the low side of the cap, it had the advantage that the output voltage is useful before the cap is charged.
typical mosfet + amp + shunt.
take care of thermal dimensioning.

Online Berni

  • Super Contributor
  • ***
  • Posts: 4911
  • Country: si
Re: Limiting initial charge current into supercap from a boost converter?
« Reply #10 on: June 21, 2021, 05:54:46 am »
Yep a boost converter will not be able to regulate current below its input voltage since it passes trough the diode.

There are also boost converters designed for providing a regulated current. They are called switching boost LED drivers. They are mostly the same as a regular boost converter except the voltage on the FB pin tends to be set fairly low so that you can directly connect it to a shunt resistor (without burning a lot of power as heat on it). You can use any boost converter chip in current regulation mode if you replace the voltage divider with a shunt resistor.

You can solve the issue with current passing trough the diode by slightly modifying the topology. The inductor in a boost regulator chip can be swapped out for a flyback transformer, or you can use a double winding inductor made for SEPIC switchers as a transformer. This decouples the input and output and gives it buck boost capability of outputting less than the input voltage.

The last time i had to simply charge a supercap i just used a string of power resistors that get controlled by a mosfet that is being driven by a voltage supervisor chip to stop charging at a certain voltage. Its not very efficient as it burns a good deal of the energy as heat, but it is simple.
 

Offline wb0gazTopic starter

  • Regular Contributor
  • *
  • Posts: 190
Re: Limiting initial charge current into supercap from a boost converter?
« Reply #11 on: June 21, 2021, 01:06:09 pm »
Hello and thank you again for the very helpful ideas!

Berni - your idea of enabling a series resistance during start-up seems like the simplest alternative if I keep the existing simple boost regulator. Other regulator/converter types I will need to learn more. For switched series resistance, I think p-fet in series pass controlled by comparator referencing to ~5V could be sufficient. As earlier posts suggest, adding work-around is probably not wise compared to replacing converter topology with appropriate type, however, sourcing simple ready-made module for this configuration may not be so easy as product documentation for ebay-sourced converter modules is very limited.

Also, I failed to include one more detail about my expected operating regime - after initial start-up (capacitor 0V until > input voltage level 4-5V), then the application current pulse will not reduce voltage to this level (4-5V) during normal operation. From start-up, voltage sequence at supercapacitor will be 0V -> 4-5V -> 12-14V -> 10-12V -> 12-14V -> ...

Thank you again.
 

Online Berni

  • Super Contributor
  • ***
  • Posts: 4911
  • Country: si
Re: Limiting initial charge current into supercap from a boost converter?
« Reply #12 on: June 21, 2021, 04:11:40 pm »
You don't really need the full documentation to hack a switching regulator IC to do something it is not supposed to do since most of them do the same thing. If you are familiar how the topology it uses works then you know what that chip is going to do. The things are mostly just an error amplifier that compares the internal reference with the voltage on the FB pin, then runs that trough a PWM oscillator that then turns the actual power transistor on and off quickly. You don't have to treat switching regulators as a unknown black box where you have to wire it up exactly according to the datasheet. Hence you can use a regulator chip to implement any topology that is similar enough.
 

Offline wb0gazTopic starter

  • Regular Contributor
  • *
  • Posts: 190
Re: Limiting initial charge current into supercap from a boost converter?
« Reply #13 on: June 21, 2021, 04:16:36 pm »
Thanks, Berni - I was getting that impression (that it's a somewhat basic function device) from looking at the ME2149 data sheet, however, I'm also gaining respect for the details involved in successful switching converter design/implementation, my need is a one-off, and this is not my main area of experience, so I'm hoping to work at the module level rather than the IC level. At least I've got a handle on the vulnerability of the design of the converter module I started with, however, when I search for "CV/CC" type boost converters, I'm not finding any yet that support a transient short-circuit load condition where current continues to be delivered.

Dave
 

Offline sandalcandal

  • Supporter
  • ****
  • Posts: 641
  • Country: au
  • MOAR POWA!
Re: Limiting initial charge current into supercap from a boost converter?
« Reply #14 on: June 21, 2021, 05:00:13 pm »
when I search for "CV/CC" type boost converters, I'm not finding any yet that support a transient short-circuit load condition where current continues to be delivered.
For the aforementioned reasons you won't get CC control, especially with short circuit for a pure boost converter. You'll need some sort of buck-boost or flyback. Does this work? https://www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20210621085745&SearchText=cc+cv+buck+boost
Disclosure: Involved in electric vehicle and energy storage system technologies
 

Offline wb0gazTopic starter

  • Regular Contributor
  • *
  • Posts: 190
Re: Limiting initial charge current into supercap from a boost converter?
« Reply #15 on: June 21, 2021, 05:11:18 pm »
That's getting me closer, however, the various items brought up from that link all seemed to have minimum input voltage of 5V (my source is nominally 4-5V).

Looking at a few "buck boost" ICs - these reflect a different arrangement than the ME2149 --- the input (source) goes into the IC, there's an inductor separately attached to the IC, and the output (sink) leaves from the IC. This suggests there's no direct path from source to sink as there was in the ME2149 implementation.

I'll focus my search on "buck boost" and see what I can dig up that's closer to the electrical parameters I have to work with.
 
The following users thanked this post: sandalcandal

Offline wb0gazTopic starter

  • Regular Contributor
  • *
  • Posts: 190
Re: Limiting initial charge current into supercap from a boost converter?
« Reply #16 on: June 22, 2021, 06:25:50 pm »
Some more searching led to a class of devices that could help in a range of cases where the charger device cannot accommodate the charge-up profile of a supercapacitor - I searched for "thermal limited load switch IC"; these devices are based on P-FET (placed in the high side connection from source to sink) with some added functions, allowing current to be limited, thermal conditions to be accommodated, and provide various control/status interfaces.

I have a couple of SEPIC dc/dc converters on the way (that accommodate short-circuit condition; do not know yet if they'll deliver current in that condition) and will explore these thermal limited load switch devices further.

Also found interesting reading courtesy of cap-xx which published this (and a few other) application notes that helped shed light on use and applications of supercapacitors:

https://www.cap-xx.com/wp-content/uploads/2015/04/1209-CAP-XX-Supercapacitors-as-Power-Buffers-Battery-Power-2012.pdf

Thanks for all of the helpful discussion!
 
The following users thanked this post: sandalcandal


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf