Can we use such PWM duty cycle as a DC to DC cellphone charger/converter?

[Willen]

If supply is 12V then may be I can get 5V by approx. 45% duty cycle. But I felt that the pulses are not axactly 5V even in 45% duty cycle but total power like 5V. Actually it is 12V pulses and that can hit delicate circuit, isn't it? Or I can make DC to DC charger (12V to 5V) for cell phone?

 

[crutschow]

Yes, your circuit will generate 12V peak voltage across the charger. To convert 12Vdc to 5Vdc with PWM you need an inductor and a flyback diode. A motor doesn't need the inductor since it has its own inductance. And you need the phone charger in parallel with the output, not in series.

How much current does the charger take?

 

[Willen]

Hi
Oh wow! Feeling nice by knowing that at least I can use this circuit by little modification! I need almost 700mA Max current for simple cell phones cherge. (If I need current booster MOSFET then I have IRFZ44 too, if it is applicable here). Will you guide me to use series inductor and flyback diode to make well DC to DC charger? Excited! (I hardly searched MC34063 but sadly unable to get, so this one is alternative for me.)

And at last, how much frequency is better? (leave it if frequency is not much more critical or leave it if circuit is fine for frequency matter)

 

[crutschow]

Below is a sim of the 555 PWM circuit with a MOSFET output driver (using the IRFZ44 you said you have), an inductor and a diode. You can see it's delivering a little over 0.7A at 5V.

Be aware that the output voltage is unregulated and will vary proportional to the input voltage and some with load. The pot U2 adjusts the PWM duty-cycle to give the desired voltage (adjust at the desired load and expected input voltage). RminLoad is added to give a minimum load when the charger isn't connected, otherwise the output will go to 12V. Note that the bottom lead, labeled "V+Load", is the positive output to the charger (and connected to the plus side of the 12V supply) and the top lead is the charger negative (common) pin. (This means you can't connect the +12V ground and the charger ground together, but that shouldn't be a problem).

The circuit runs at a little over 100kHz PWM frequency which is probably about as high as you want to go with that circuit.

The diode can be any 1A or larger Schottky type diode.

The inductor must be able to carry 1A current without saturating, otherwise the MOSFET power dissipation will go up and you can blow the MOSFET. It should be 200μH minimum or larger.

Edit: Below is a modified circuit that reduces the inductor current.

 

[crutschow]

As an alternate, below is a sim of a relatively simple Hysteretic (bang-bang) type switch-mode converter which automatically sets the PWM to give a regulated 5V output independent of input voltage and load variations. It has a somewhat higher output ripple then other types of switch-mode supplies (about 50mV for the circuit shown at 700mA) but this should be no problem for a cell-phone charger.

It uses a commonly available comparator as the control circuit (LM339 or LM393).

The one possible disadvantage of this circuit as compared to the 555 circuit is that it requires a P-MOSFET rather than an N-MOSFET for the switching transistor. The advantage of a Hysteretic converter is that it doesn't require compensation in the feedback loop for stability as the more common types of switch-mode supplies do.

The minimum required inductor size is only 20μH, but due to the nature of the circuit operation it must handle a peak of about 4A without saturating.

 

[dr pepper]

I have built similar circuits to the 555 one, with and without feedback and some with a transformer rather than an inductor for isolation, its a veritable circuit.
Not having feedback has the advantage that its less likely to get full 12v supply voltage to the load.
Your original circuit might blow up the phone, you need a inductor/diode.

 

[audioguru]

The charging circuit in the phone does not continuously draw 720mA. The current drops fairly low as the battery nears a full charge. Then the output voltage is limited only by Rmin that might blow up the charging circuit.
Instead, why not use the 5V power supply ("charger") that came with the phone?

 

[Willen]

I am interested to 'Hysteretic Converter' by crutschow because most of the time here being power failure (loadscheding actually). I was was excited to Crutschow's 2nd comparator based circuit and also was searching P-channel MOSFET. As you (AG) said- will it be dangerous?

 

[crutschow]

My first circuit's voltage would increase some under no load, since it operates open-loop, but the Hysteretic Converter would not as it's closed-loop. Its voltage would only increase if there were a circuit failure of some type.

Any P-Channel MOSFET with a voltage rating of 20V or greater, an ON resistance of 0.1Ω or less, and a gate charge of no more than about 20nC should work.

What is the 12V source you would use?

 

[audioguru]

My reliable electricity is provided by nuclear and a few huge waterfalls. The coal-burning steam engine powered electricity generating plants have all been closed due to their air pollution. Some are fuelled by natural gas. My government is giving grants to people who install solar power and feed it into the grid. There are very few huge windmills here but there are many of them used in some countries.

 

[Willen]

Windmills looks so interesting! Our country has only Hydropowers (waterfalls) as a electricity source. We have total 83,000 MegaWatts 'probal' electricity capacity but we just being able to produce around 500 MegaWatts till now. So facing heavy loadscheding, mix 18 hours per day in spring.

Hi crustschow,
I am going to use 12V lead acid battery. In the Hysteretic Converter, I think a part LT1634-5 is a 5V (5.1V available) standard zener diode and near 150nF is not more critical (maybe I can use 100nF or 220nF). And maybe 20uH is not MORE critical. I think 30 turns 24SWG wire on 1.5cm (diameter) ferrite ring is fine.

I got lot more comparators from few helpful Americans, I hope I will find P-channel mosfets in my local shop (probably P-channel is less accessible than N-channel).

 

[dr pepper]

The inductor is an important component, you can go a bit higher on inductance but not lower, also the dc current rating has to be at least twice the input current to the circuit or it will saturate and overheat.

 

[crutschow]

..............................
In the Hysteretic Converter, I think a part LT1634-5 is a 5V (5.1V available) standard zener diode and near 150nF is not more critical (maybe I can use 100nF or 220nF). And maybe 20uH is not MORE critical. I think 30 turns 24SWG wire on 1.5cm (diameter) ferrite ring is fine.
..................................

LT1634-5 is an IC that acts as a near ideal 5.0V zener. You could use a standard zener but the standard closest value is 5.1V with a higher tolerance than the LT1634 so you output voltage will be likely some different than 5.0V, but that probably is tolerable for the charger.

Increase C3 to 1μF as that slows down the startup surge and reduces the peak current through the inductor.

As dr pepper noted, the saturation current of the inductor is critical. It needs to be about 3A (with R3 = 20kΩ and C3 = 1μF). Do you have an oscilloscope so you can check the saturation point and inductance value of the inductor?

 

[Willen]

I can apply TL431 instead of LT1634-5 for precise 5V however 5.1V is tolerable as you said. I am a just hobbyist and always want to make such things madly! I know how transistor saturates but I haven't read how inductor does. But feeling exciting to read its basic and will read soon!

 

[crutschow]

I can apply TL431 instead of LT1634-5 for precise 5V however 5.1V is tolerable as you said. I am a just hobbyist and always want to make such things madly! I know how transistor saturates but I haven't read how inductor does. But feeling exciting to read its basic and will read soon!

A TL431 should work fine to generate 5V.
But I doubt that the charger would mind 5.1V.

When an inductor saturates as determined by the ampere turns and core material (meaning the core can accept no more flux) then the inductance drops to near zero and the inductor now acts like a piece of wire rather than an inductor. This can cause large currents through the transistor if the circuit depends upon the inductance to limit the current (as the Hysteretic Converter does).

 

[Willen]

Hi crutschow,

In local poor store, I found almost 30 types of N-channel MOSFETs but hardly found a P-channel MOSFET which is IRF9620. I have not seen its detail datasheet but according to its quick note it has Rds(on)=1.5 ohms, 200V and -3.5A. You were saying that I need 0.1 ohms MOSFET.

Here, I expect that If I used 5.1V ZD as reference then this 1.5 ohms ON resistance will maintain (decrease) this 0.1V and I will get around 5V. Can I think like this way?

 

[audioguru]

There are thousands of modern low Rds(on) P-channel Mosfets available. The IRF9620 has a fairly old design and has a fairly high voltage rating which combine to make its on-resistance high.
Its maximum on-resistance is 1.5 ohms when it has 10V for Vgs.

 

[crutschow]

...................................
Here, I expect that If I used 5.1V ZD as reference then this 1.5 ohms ON resistance will maintain (decrease) this 0.1V and I will get around 5V. Can I think like this way?

No. The output voltage is determined by the reference voltage compared to the output voltage by the negative feedback and is not affected by the MOSFET ON resistance. The main effect of the higher ON resistance is to increase the power dissipated in the MOSFET and reduce the converter efficiency.

That being said, 1.5 ohms of ON resistance is not prohibitive in this case, generating a little less than a watt of dissipation in the MOSFET at 700mA output current.

Edit: Further simulations indicate the MOSFET dissipation would be about 1.6W with a 1.5 ohm ON resistance which means the MOSFET would get fairly hot without a heat-sink.

 

[rumpfy]

Hi Willen,
just to take a different direction, Many mobile phones have a charger which plugs into the cigarette lighter. In australia, i have seen many many of these chargers in the scrap bins as people upgrade to a new phone. Australia has 22 million people and 30 million mobile ohones.
I notice that in EVERY case I have seen, the charger uses a '34063' IC to do all the required functions. It is a PWM type of device but includes it's own feedback controller and the switching transistor.
I have recently adapted an old charger to suit an old Nokia mobile phone.
The phone itself has its own inbuilt charge controller for the NiMH battery. The charger itself just supplies a constant DC voltage and the phone adjusts the pulse width of the charging current. So you dont have to worry about overcharging the phone battery.
I have used the 34063 for many power supplies over the years. The net has a very good range of Applications Bulletins.
The 34063 has a peak current rating of 1.6 amp so it will charge phone batteries at the 0.8 amp level quite happily. The 34063 is an 8 pin IC.
Hope this helps and doesnt confuse you!.

 

[Willen]

The sentence made me amazing and confusing what you were saying-

Many mobile phones have a charger which plugs into the cigarette lighter.

Hi rumpfy,
Your description is so nice for me. Once I also found a 34063 chip on my uncle's Chinese simple DC to DC charger. I felt the necessary of DC to DC converter chip after seeing the charger. After then I searched the 34063 on local market but failed. Datasheet of 34063 had made me excited.

Due to its unavailability, I started to think about alternative way. Crutschow suggested so nice converter by comparator and P channel mosfet but unluckily I found one which was not suitable. I found almost 30 types of N-channel MOSFETs anyway.