How To Improve Boost DC/DC Converter?

[devonsc]

boost dc

Hi there,

I'm working on a boost DC/DC converter to charge a 12V sealed lead acid battery from a solar panel at the moment. Basically, I've replaced the BJT with MOSFET and the diode with a Schottky Diode. Apart from all this, how do I further increase the efficiency of my DC/DC converter?

I will be feeding the MOSFET with a PWM signal from a PIC Microcontroller. Is there anything that I can do with the PWM signal to obtain a maximum power point at the output of the boost DC/DC converter? Should I control the PWM signal by getting some feedback from the battery to control the step-up voltage level? By doing so, am I improving the efficiency of the boost DC/DC converter? If yes, mind to suggest on how can I get it done? Please?

Any advice? Help really needed...Thanks in advance...

 

[nicleo]

solar dc/dc up converter

Do you have the following ebook?


Hope it will help you.

You might consider ZVS or ZCS to improve the efficiency of your boost converter.

 

[devonsc]

dc-dc converter basics boost

Oh no! I do not have sufficient points to download the notes you recommended...Any way of accessing the E-book you suggested? Need help here, rushing to complete my project...Please?

Advice needed...

 

[Black Jack]

ogd-12v drawing

Oh no! I do not have sufficient points to download the notes you recommended...Any way of accessing the E-book you suggested? Need help here, rushing to complete my project...Please?

Advice needed...

See National Semiconductor`s integral switch.
www.national.com

 

[olemp]

boost dc dc converter,max duty

feedback relevant with your battary charge strategy....
it doesn't effect the efficiency of your converter....

use a mosfet which has small rds on, make small enough the resistance of boost enductance(increasing thickness of the wire )...schottky good choice..and your switching frequency should be 50khz((+)or(-)10khz)
in my opinion those are enough...
for further efficiency you should use zero voltage or zero current switching and change the schottky with mosfet...but those adds complexity the your design and it improves efficiency approximately %5-%10.if this is important for you use ZVS or ZCS topology.........

 

[platonas]

irl2703 app note

Hello devonsc
Just few suggestions to improve the efficiency of any switching regulator.
1. Use high quality, low ESR electrolitic capacitors,
2. Use a Mosfet with very low ON resistance,
3. The coil/transformer that you use, can be wound with more than one wires in parallel in order to form the desired guage. With this I mean, instead of using a single 1mm wire, you can use 5 wires of 0,2mm and wound all together, like a multicore wire. This method, decreases the resistance of the wire because at high frequencies you have the so called skin effect. Current travels at the skin of the conductor, the surface.
The surface of the 5 wires of 0,2mm is more than that of the single 1mm wire.
I always talk about insulated wire, the one that is proper for the coils.

Platonas

 

[devonsc]

site:www.edaboard.com current limit boost

use a mosfet which has small rds on, make small enough the resistance of boost enductance(increasing thickness of the wire )...schottky good choice..and your switching frequency should be 50khz((+)or(-)10khz)

Thanks for correcting me, initially, I was just trying to use 4kHz. Thanks a lot. By the way, what are the consequences of using a fast-switching diode instead of a schottky diode. Generally, between this two diodes, which will perform better in increasing the efficiency? Need advice...

Thanks in advance...

 

[devonsc]

ntr4503

1. Use high quality, low ESR electrolitic capacitors,
2. Use a Mosfet with very low ON resistance,

Thanks...

3. The coil/transformer that you use, can be wound with more than one wires in parallel in order to form the desired guage. With this I mean, instead of using a single 1mm wire, you can use 5 wires of 0,2mm and wound all together, like a multicore wire. This method, decreases the resistance of the wire because at high frequencies you have the so called skin effect. Current travels at the skin of the conductor, the surface.
The surface of the 5 wires of 0,2mm is more than that of the single 1mm wire.
I always talk about insulated wire, the one that is proper for the coils.

You mean I should build my own inductor? Is that going to be real difficult

By the way, building the boost DC/DC converter is part of my project. I will need to learn how to develop an AC/DC converter after completing this.

Help really needed...Thanks a lot

 

[devonsc]

improve solar panel efficiency dc dc converter

Real sorry if this is nonsense, I have very limited electronic knowledge. Regarding the switching of a boost DC/DC converter, does it mean that the faster the switching time, the better it is?

 

[nicleo]

boost dc/dc max duty cycle

Real sorry if this is nonsense, I have very limited electronic knowledge. Regarding the switching of a boost DC/DC converter, does it mean that the faster the switching time, the better it is?

If the switching frequency is higher, the size of the DC/DC converter will be smaller.

 

[platonas]

very simple dc-dc boost circuit

If you cannot build your own inductor is ok, just forget this part.
As far as the diode is conserned, use a fast one with the lower forward voltage drop.

 

[gevy]

losses converter switch boost quiescent

devonse wrote:

Oh no! I do not have sufficient points to download ...

See link Free mirror

I attach the useful chapter from the book

 

[devonsc]

ntr4503 pdf

Hi there,

I would like to ask about something that I've miss-out in developing the DC/DC converter. Recently, I come across something called operating the converter under continuos mode or discontinuos mode. But I don't really get what is/are the differences between these two. Do you guys mind to brief me on this matter? Please?

Thanks in advance...

 

[nicleo]

dc dc converter solar panel

I would like to ask about something that I've miss-out in developing the DC/DC converter. Recently, I come across something called operating the converter under continuos mode or discontinuos mode. But I don't really get what is/are the differences between these two. Do you guys mind to brief me on this matter? Please?

When DC/DC converter, e.g. buck converter, works in 'continuous' mode, the inductor current is continuous (see figure (1)). When DC/DC converter works in 'discontinuous' mode, the inductor is discontinuous, meaning the inductor current is zero at certain portion of each cycle (see figure (2)).

We can see the effect of discontinuous current on the voltage ratio of the converter in figure (4). As seen in the figure, once the output current is high enough, the voltage ratio depends only on the duty ratio "d". At low currents the discontinuous operation tends to increase the output voltage of the converter towards Vin.

Hope it helps.

Regards.

 

[devonsc]

pgate+switching voltage regulator+edaboard

Thanks a lot, everyone...

I'm not sure if I'm allow to ask this here...about the term Reverse Recovery Time, I found this term in the diode's datasheet. Was wondering if this one of the parameters that I should take into consideration while selecting my diode? Does this parameter have anything to do with the switching time?

Initially, I selected BAT48 as my diode, but there isn't any Reverse Recovery Time parameter listed in its datasheet. I've selected BAT48 due to the low forward voltage. Am I in the right track?

By the way, I was trying to make some comparisons between 1N4448 diode and BAT48.

(My progress is real slow in developing this DC/DC converter, I will post my progress so far later on)

Guide really needed....Really mean it when I thank you guys...Thank you.

 

[nicleo]

irms mosfet tapped-boost

I'm not sure if I'm allow to ask this here...about the term Reverse Recovery Time, I found this term in the diode's datasheet. Was wondering if this one of the parameters that I should take into consideration while selecting my diode? Does this parameter have anything to do with the switching time?

When the switching frequency is higher (20kHz), standard power diode, e.g. 1N1401, can not be used anymore because of the 'reverse recovery' problem. In higher switching frequency application, you need to choose fast recovery diode.

BAT48 datasheet mentions that it's designed for fast switching application. But I could not find any info (regarding reverse recovery time) in 1N4448 datasheet. You can use signal generator to test the diodes at high frequency. If the waveform is not distorted, then they are OK for high switching application.

 

[olemp]

low power solar panel,7.5v 150ma

I've selected BAT48 due to the low forward voltage. Am I in the right track?

if this is for power diode this is not appropriate.. for 4 amper you can use two parallel 1N5822...
http://www.fairchildsemi.com/ds/1N/1N5822.pdf

for reverse recovery time look at this:
http://www.powerdesigners.com/InfoWeb/design_center/articles/Diodes/diodes.shtm

if you are using high frequency diode and using it in low currents (<8-10 amper(also low voltage<40-50v))
don't think about it ... it is not problem....but above these ratings you have to use R-C snubber across the diode....
if you want to learn extremely theorical information it takes a lot of time.learn by appliying.. just do it and test whether it meets your requiretments or not.

 

[electronrancher]

how to improve boost converter efficiency

first things first. lower switching frequency is better - always! people like converters to work >20kHz so you can't hear them whine under heavy load, but for your application it doesn't really matter.

second - either fast switching, or schottky is OK. You want low resistance, low forward voltage, and fast recovery. There is loss when the diode needs to stop conducting and start blocking - faster recovery means less loss.

third - operate in CCM if you can. This means choose an inductor large enough to keep the ripple current above 0 at it's lowest point. for example, with 1amp load you should try for maybe 400-500mA ripple. If you sometimes operate in a no-load condition you can't avoid DCM so don't worry about no-load conditions.


I would disregard discussions about using high switching frequency in order to use tiny inductors. The reason people use high freuquency is to get the inductor small enough (0.6uH) to fit in a cellphone or something. You don't care about this, you're charging a huge lead acid - your application will not be small anyway. You're looking at an inductor in the 10-20uH range, maybe bigger for 4kHz.


In order to implement MPPT (which I think is possible with a PIC - no problem) you will measure the solar cell output voltage and current. Keep increasing current until the product of Vout*Iout begins to drop. That is MPP. Back off the current a little and you will see it peak again. Back off too much and it will droop. Best MPP keeps sweeping back and forth across the range, reducing load when a cloud comes by and chokes down the cell, increasing load when it is very bright. making this adjustment once or twice a second is fine.

MPPT looks for highest efficiency of the CELL, not the DC-DC Converter. DC-DC efficiency is governed by the things discussed above, like diode drop and switch on resistance - If your MPP is 10Watts and you have 80% efficient DC-DC, you get 8W with which to charge the battery. If you don't use MPPT, maybe you drive the cell too hard and only get 6W. Again with 80% efficient DC-DC this only gives 4.8W to charge the battery.


One note - you will need a driver IC in order to switch that mosfet. PIC cannot do it alone. Peak current into the gate should be about 1A in order to get fast switching.

 

[devonsc]

reducing esr at the output of dc boost

first things first. lower switching frequency is better - always! people like converters to work >20kHz so you can't hear them whine under heavy load, but for your application it doesn't really matter.

Hi. Would like to reconfirm with you, about the frequency of my PWM as I first intend to use high frequency, approximately 50kHz due to some certain reading materials that I've come across. Meaning, I should use less than 20kHz and still will be able to perform well? Is the following application alright? Thanks in advance.

> Crystal for my PIC microcontroller = 4Mhz
> Frequency of my PWM = 10kHz
> With reference to the datasheet of my PIC, I will obtain a resolution of 8.64 bits

OR

> Crystal for my PIC microcontroller = 4Mhz
> With reference to the datasheet of my PIC, I will obtain a resolution of 8 bits
> Frequency obtain at this level = 15.625kHz

Sorry, but if you don't mind, do you mind to recommend which approach is better? The first or the second?

One note - you will need a driver IC in order to switch that mosfet. PIC cannot do it alone. Peak current into the gate should be about 1A in order to get fast switching.

If I chose to use logic-level MOSFET, do I still need MOSFET driver?

Advice really needed. Thank you very much.

 

[electronrancher]

either 10kHz or 15kHz should be fine, but keep in mind that it will emit a high-pitched whine like an old TV under heavy load (and maybe very light load)

to me there are 3 main regions of operation.
industrial: 5kHz motor drivers to 50kHz high-power supplies (>1000W)

commercial: 100kHz to 1000kHz, desktop/PC/rack mount power. (5-500W)

portable: 1.2MHz and up. cell phones, ADSL, etc. (0.1-10W)

so 10kHz or 15kHz is slow, efficient power. very good for something like sunlight where it changes over seconds or minutes. if you get audiobanding to the point that the noise is annoying, increase to 25 or 30kHz.

8 bits of resolution means smallest duty cycle step is 0.4%. Should be plenty. What you want to avoid is something like 5% step where in order to get 51% duty your converter must give two 50% pulses and one 55% pulse.. although this doesn't really cause ANY problem, it looks like jitter/instability on the scope so people get scared.

probably a maximum step size of 1% duty is even ok, meaning 128 LSB or 7 bits is if you want to try to squeeze 20kHz out of it.

and last - yes, you need a driver, or a bunch of hex inverters in parallel (I MEAN LOTS!) even if you use logic level mosfet. here's why.

Assume the gate of one of these big mosfets is about 2nF. We can use I=C*dv/dt to find out current drive needed.

For 50ns risetime for 0-5v step on the gate requires 200mA peak current, 10ns rise needs 1A peak current. This is way too much for the little PIC output inverter, and it is the reason I see many PIC projects have problems. Risetime slower than 50ns means you start to lose efficiency, so just keep it fast and don't worry about mosfet in resistive region.

Now that I think about it, maybe a bunch of hex inverters is OK. Output of PIC drives 3 inverters in parallel, and these guys drive 8 inverters in parallel that switch the gate. 8x at 50mA should be enough for 50nS risetime..