Series Regulation

[blueroomelectronics]

This has got to be a school project.

 

[kchriste]

Can I keep the battery voltage with in limits with out wasting power?

The advantage of a shunt regulator is that it is really simple. The disadvantage is that it will dissipate more power as heat than a series regulator when the battery is fully charged. Both will charge the battery just as fast.
Once the battery is charged, there is no getting away from wasting power, no matter which method you use.
If you want to avoid heat build up, you need to use PWM in your series regulator.

 

[premkumar9]

If you want to avoid heat build up, you need to use PWM in your series regulator.

I didn't want to waste the power from solar panels.

 

[premkumar9]

This has got to be a school project.

What do you mean? Too simple?

 

[Thunderchild]

I can't see what all the fuss is about, you want to run leds from a variavble power source ? just use a constant current generating circuit, if you worried about efficienzy then consider a switch mode regulator

 

[premkumar9]

The advantage of a shunt regulator is that it is really simple. The disadvantage is that it will dissipate more power as heat than a series regulator when the battery is fully charged. Both will charge the battery just as fast.
Once the battery is charged, there is no getting away from wasting power, no matter which method you use.
If you want to avoid heat build up, you need to use PWM in your series regulator.

Refer the schematic below
**broken link removed**
My load is 600mA at 12 V supply. The FET is driven from a microcontroller. This MC doesn't have a PWM output. The solution came to my mind is:
1. Sense the current through the load.
2.Convert to binary and feed to the MC .(I think three bits are enough for my requirement)
3.MC will change the duty cycle to the driver input.
Request your comments and suggestions on how to implement.

 

[ericgibbs]

Refer the schematic below

My load is 600mA at 12 V supply. The FET is driven from a microcontroller. This MC doesn't have a PWM output. The solution came to my mind is:
1. Sense the current through the load.
2.Convert to binary and feed to the MC .(I think three bits are enough for my requirement)
3.MC will change the duty cycle to the driver input.
Request your comments and suggestions on how to implement.

I would suggest the same solution as 'TChild' whats the problem with a 600mA constant current source.?

What configuration are you using for the LED's [series/parallel] and their voltage drop,?

 

[premkumar9]

I would suggest the same solution as 'TChild' whats the problem with a 600mA constant current source.?

What configuration are you using for the LED's [series/parallel] and their voltage drop,?

Problem with constant current source (I think, you mean shunt regulation) is the poor efficiency, especially at higher input voltages (initial cost per watt for solar energy is high)

I use series parallel combination.
Using different colors of LEDs. Voltage drop 2V, 3.5V etc for different types.
At 12v supply, current is adjusted to be around 20mA through each LED.

 

[ericgibbs]

Problem with constant current source (I think, you mean shunt regulation) is the poor efficiency, especially at higher input voltages (initial cost per watt for solar energy is high)

I use series parallel combination.
Using different colors of LEDs. Voltage drop 2V, 3.5V etc for different types.
At 12v supply, current is adjusted to be around 20mA through each LED.

hi,
Constant current is not shunt regulation.

 

[blueroomelectronics]

Are there no electronics that need power?

 

[premkumar9]

hi,
Constant current is not shunt regulation.

Can you please explain little more about that. Will it take care of my efficiency problem?

 

[premkumar9]

hi,
Constant current is not shunt regulation.

I went through constant current source. i think, efficiency problem will remain in that case also. Right?
I think the only solution is switching regulation. Do you agree?

 

[ericgibbs]

I went through constant current source. i think, efficiency problem will remain in that case also. Right?
I think the only solution is switching regulation. Do you agree?

Have a look at this basic CC circuit.

 

[Thunderchild]

at the end of the day ANY current and/or voltage regulation wastes current whichever type of linear regulating circuit you use, the only way to obtain maximum efficienzy is to use some sort of switch mode regulator

 

[premkumar9]

Have a look at this basic CC circuit.

Here also I think it is again like a series regulator .My load will take 600mA current if 12V is applied across it. So here the CC regulator will adjust the voltage in such a way as to keep the load current 600mA. Then
1.Min input has to be 14.5V (you indicated 2.5 V difference required between i/p and o/p)
2.If my input is 15.5, then 3.5 x 0.6 W will be dissipated by the regulator.
(In this case I have already made the LED lamps with series parallel combination with one series resistance with each branch. So a constant voltage is enough for me. If I remove the resistance and connect LEDs only, then the current sharing of branches-even if total current is 600mA- will not be same since my LED V/I characteristics may not be exact.
Again I may have to conclude that switching regulation alone will meet my requirements

For your comments please.

 

[premkumar9]

at the end of the day ANY current and/or voltage regulation wastes current whichever type of linear regulating circuit you use, the only way to obtain maximum efficienzy is to use some sort of switch mode regulator

Just now I also replied to ericgibs in the same line. So you may pls look the diagram and remarks and give your comments on how to sense the current/voltage and generate 3 bit information for my MC
(In fact i have already completed the work and installed the unit in the site. But one mistake I did was I made an assumption that the DC from charge control unit is regulated 12V- communication gap between me and supplier. Now i have to work out a solution which can be added to that. That is why I planned the solution with minimum disturbance to the existing circuit.)

 

[premkumar9]

OK. Now it is 10.45 PM here. I will check the mails again at 6AM tomorrow. Bye

 

[audioguru]

With the Mosfet driven by PWM then the current-limiting is provided by the battery or the Mosfet. With the duty-cycle reduced then the current in the LEDs must be much higher to provide the same brightness as with DC.
So the PWM circuit is not more efficient than a linear constant current source.

 

[premkumar9]

With the Mosfet driven by PWM then the current-limiting is provided by the battery or the Mosfet.

Let the input votage be 15.5V. The PWM will reduce the duty cycle. But peak current will be higher as the voltage applied is higher (FET acts as switch and battery has high current capacity). So I don't think battery or FET will not limit current. Right?

With the duty-cycle reduced then the current in the LEDs must be much higher to provide the same brightness as with DC.
So the PWM circuit is not more efficient than a linear constant current source.

It is true that higher current to be driven when duty cycle is less. But in the case of linear constant current source voltge is dropped by the series device and hence power loss. But in the case of PWM, the FET acts as a switch and hence efficiency is very high. So almost full power is supplied to the load.

 

[audioguru]

But in the case of PWM, the FET acts as a switch and hence efficiency is very high. So almost full power is supplied to the load.

Then the LEDs instantly burn out because the peak current is too high.

Of course the battery limits the current. It and its connecting wires have resistance.
The mosfet also has on-resistance that limits the current.
These current-limiting resistances either are too low so the current is too high and the LEDs burn out, or the resistances waste a lot of battery power like a linear regulator.

PWM reduces the heating of the Mosfet driver if its on-resistance is low enough. An external current-limiting device is needed and it wates power by making heat. Efficiency is not any better than a linear regulator and might be worse.

A buck converter increases efficiency.