designing a charge controller for hybrid system

[diksha.k]

actually everything seemed perfect but, when i really realized that though i was getting a step down in voltage but,the mosfet where not getting hot at all with battery as load then,i checked for Vgs voltage=8.3V,Vds voltage=0V
If MOSFET was conducting then there had to be drop around 3V.
Now the solar voltage also reduced with increasing duty cycle But the thing is Vds=0v only there has to be atleast some drop.

 

[diksha.k]

You should be able to load the solar panel right down to battery voltage at maximum full duty cycle.

No this was not observed

At minimum duty cycle the solar panel voltage should be able to go right up to the full open circuit voltage.

This could easily be observed.

 

[Warpspeed]

With the mosfet hard on all the time, there is only the mosfet and inductor in series between the solar panel and the battery.
Where is this voltage drop ?

 

[diksha.k]

With the mosfet hard on all the time, there is only the mosfet and inductor in series between the solar panel and the battery.
Where is this voltage drop ?

I meant turn on voltage drop of MOSFET i.e Vds will be 2-3V right?

 

[Warpspeed]

I meant turn on voltage drop of MOSFET i.e Vds will be 2-3V right?

I hope its not that high.
What mosfet are you using ?

Even with something very ordinary like a single IRF540 has an on resistance of .055 ohms. Even at ten amps, that is only 0.55 volts drop.
Mosfets with only .005 ohms on resistance are out there, and they are not expensive.

 

[diksha.k]

I hope its not that high.
What mosfet are you using ?
.

IRFZ44N.
But,As i earlier told you that:
with 99% duty cycle panel voltage from 21V drops to 18V whereas battery voltage=12.2V.
And with 100% duty cycle panel voltage remains at Voc . with 1%/0%duty cycle panel voltage again remains at Voc.

- - - Updated - - -

- - - Updated - - -

mosfet driver i am using is:
https://drive.google.com/file/d/0B9S6uuNt7m4OUlNjaTFIUDhoTGc/view?usp=sharing
this is working fine i have tested it.

 

[Warpspeed]

At 99% duty cycle there should be much less than 200mV drop.
Rds on of IRFZ44 is quoted as 17.5milliohms.

Are you measuring these output voltages with a load on the output of the buck regulator?

With a battery connected, you should be able to adjust the solar panel voltage right down to battery voltage within 100 to 200mV at full maximum duty cycle.

 

[diksha.k]

At 99% duty cycle there should be much less than 200mV drop.
Rds on of IRFZ44 is quoted as 17.5milliohms.

Are you measuring these output voltages with a load on the output of the buck regulator?

Yes i have connected the battery as load.

 

[Warpspeed]

Try connecting a 9v battery directly between gate and source.
That should turn it on 100%, and see what voltage you measure with that.

Something is definitely not quite right somewhere.

 

[diksha.k]

Try connecting a 9v battery directly between gate and source.
That should turn it on 100%, and see what voltage you measure with that.

Something is definitely not quite right somewhere.

yes i will let you know about it.
for Flywheel diode i am using a general purpose diode (IN4008) of higher rating though, most of them use schotkky diode. Will using schotkky diode make difference?

 

[Warpspeed]

A general purpose rectifier diode will be far too slow, and it needs to have a higher current rating than one amp.

There are plenty of alternative shottky diodes, but something like a MBR1645 might be about right for this.
http://pdf.datasheetcatalog.com/datasheet/motorola/MBR1645.pdf

A shottky diode will certainly improve efficiency around mid duty cycle by having a lower forward voltage drop. But up near 100 % duty cycle, the flywheel diode hardly ever turns on, so its not the root cause of the large voltage drop problem.

 

[diksha.k]

Try connecting a 9v battery directly between gate and source.
That should turn it on 100%, and see what voltage you measure with that.

Something is definitely not quite right somewhere.

It worked fine when i tried as you suggested.
Now coming to the main probelm:
it seems the microcontroller i am using isn't right.
I am using arduino board for programing.
The problem is that:
when i make duty cycle=99% the voltage of solar panel is got down to almost battery voltage. But, on 100% duty cycle it's like in open circuit.
When duty cycle=0% again solar panel voltage got down to battery voltage(weird). And when duty cycle=2% panel voltage=open circuit voltage.

 

[Warpspeed]

It sounds like all the hardware is working fine, but something is wrong with your duty cycle program in software.

 

[diksha.k]

There was some software problem.
Now results are:
when i rigged up everything with 100% duty cycle the o/p of panel dropped to 16.8V and battery voltage raised to 14.4V.
Is this ok? I also checked disconnecting the arduino same results.
Changed various mosfets again same results.
changed inductor value(i know at 100% duty cycle it acts as short) same results.
changed capacitor same results.

- - - Updated - - -

And i found that for 17V panel vmpp=14.85

 

[Warpspeed]

Great news fixing the software.
One big step closer.

Where is all that voltage being dropped ?
Something must be getting fairly warm, and that should be a clue.

 

[diksha.k]

Great news fixing the software.
One big step closer.

Where is all that voltage being dropped ?
Something must be getting fairly warm, and that should be a clue.

I measured voltage across flywheel diode(or inductor+capacitor with battery across capacitor) it was around 14.45V.
And indeed MOSFET was getting warmer but, when i measured the voltage across the drain-source it was around 0.34V falling short of:16.8-14.45-0.34=2.01V.

 

[FvM]

I'll prefer to believe the validity of Kirchhoffs voltage law https://en.wikipedia.org/wiki/Kirchhoff's_circuit_laws#Kirchhoff.27s_voltage_law_.28KVL.29

Possible explanations
1. instrument defective
2. instrument confused by switching voltages
3. wrong voltage nodes probed (did you close the loop correctly?)

I remember a recent thread about "unexplainable" behavior of a voltage regulator. Turned out as an empty multimeter battery... Edaboard members had a hard time to explain the results though :smile:

 

[diksha.k]

I'll prefer to believe the validity of Kirchhoffs voltage law https://en.wikipedia.org/wiki/Kirchhoff's_circuit_laws#Kirchhoff.27s_voltage_law_.28KVL.29

Possible explanations
1. instrument defective
2. instrument confused by switching voltages
3. wrong voltage nodes probed (did you close the loop correctly?)

I remember a recent thread about "unexplainable" behavior of a voltage regulator. Turned out as an empty multimeter battery... Edaboard members had a hard time to explain the results though :smile:

I have changed multimeter and checked.
I am performing on breadboard.
the wire for connecting negative rail of the load to negative of solar to close the circuit has a drop of 1.5V when mosfet is completely turned on(100% duty cycle)

 

[FvM]

the wire for connecting negative rail of the load to negative of solar to close the circuit has a drop of 1.5V when mosfet is completely turned on

That's a huge amount for a 12V system. But in any case it has to be considered in the loop summation, the Kirchhoff rules are explicite in this regard I believe. Does it mean that "2.01 V" difference are shrunk to actual 0.5 V?

 

[diksha.k]

Does it mean that "2.01 V" difference are shrunk to actual 0.5 V?

Actually the results are varying with +/-0.5V initially on 100% duty cycle i got around 2V difference then after doing once more time i found 1.5V difference.
This 1.5V/2V seems to be dropping along the rail connecting the negative of load &source.