Constant Current Source?

[thecritic]

Please have a Look at the following Circuit (Micro-Chip Application Note). The circuit is supposed to provide constant charger current to the battery
**broken link removed**
Ignore R9, R8 and A2D for now. They are for voltage monitoring. So, to check the concept of constant current flow, I made the following circuit using Stand-Alone Comparator

**broken link removed**

I was expecting to have the bulb glow only half (Average Voltage across it will be only 6V). And It glowed only half as brightly, just as expected.

But I was also expecting that the comparator would be bang-bang controlling the FET so, it wouldn't heat up. But In my case, the FET gets heated like hell. I want it to run cool. How do I do it? Or, the Micro-chip's design is just the crappy linear Constant Current Regulator?

 

[geko]

How do I do it? Or, the Micro-chip's design is just the crappy linear
Constant Current Regulator?

No, it's uses a buck switching regulator and L1/D2 are a very important part of that circuit as the description in the AppNote explains.
https://www.electro-tech-online.com/custompdfs/2011/05/00626a.pdf

You can't just chop a few components out and expect it to work.

 

[BrownOut]

The circuit is linear. OK, I didn't see geko's post, so I won't repeat.

 

[PICMICRO]

No, it's uses a buck switching regulator and L1/D2 are a very important part of that circuit as the description in the AppNote explains.
https://www.electro-tech-online.com/custompdfs/2011/05/00626a-1.pdf

You can't just chop a few components out and expect it to work.

My bad!. Sorry for the negligence!
So, what value of inductor/capacitor do you suggest for my test circuit so that the FET would run cool! ?

Also, I think I hear somewhere about methods of using Transformer as Inductor (I have a small 15VA 220/3 V Transformer)
Any Idea about it?

I'm going to sleep now (Night here). Meet you tommorow.

 

[PICMICRO]

The circuit is linear.

You meant mine circuit was linear (not microchip). Don't you?

 

[Mr RB]

Have you considered using two LM317 regulators? I have built a number of battery chargers over the years with that system.

The first LM317 is the constant current source, the second one is the voltage regulator. You get the same performance as your PIC version but both LM317 have added banefits of over-current limiting and thermal shutdown protection.

 

[thecritic]

I want charging current of around 15A at 12 Volts. I Think I would be wasting lot of money and energy If I go to LM317's route.
I will experiment with Inductors and Capacitors in the Test circuit now.

 

[thecritic]

Update

Here is the update to my 'test' project.
I made the following circuit. The supply Voltage was 11.9 Volts.
**broken link removed**
At first I opened the jumper JP1.
The voltage at B was 11.9 Volts. Initially the POT was all way up and the Voltage at C was also 11.9 Volts. I then Decrease the Voltage At C from 11.9 volts.
I was hoping the the output (A) would swing to High As soon as the C (I-) was below B (I+). But, in my case, I had to decrease Voltage at A down to 10.9 volts before the output switched states. Moreover, I would get intermediate output voltage (such as 5,6 Volts) by carefully trimming the POT to around the critical point. Is this normal???

Any-way I Closed the jumper then. By manipulating the POT I could easily make the Bulb glow half brightly (i.e make the Current half of the rated current) or make it glow at any brightness. i.e. I could control the current. But, the FET would heat like a frying pan! Please help me run the FET cool?

Here is the transformer (220/3 V) whose secondary I used for the inductor. So, I am not sure whats the inductance I used.
**broken link removed**

 

[4pyros]

1st off you cannt just throw any inductor in there and expect it to work. Page 5 of the app note explains how to determin the component values you need.
2nd you must put that FET on a heat sink no mater what!!!!

Andy

 

[thecritic]

Hi, 4pyros. Thanks
Without the regard to the FET and the other circuit, Do you think the behavior of the Comparator Circuit I described in the first paragraph in previous post fine? I am also in the doubt that my Comparator Wasn't good enough!

I was in the thought that, the Inductor can be any value greater than the specified minimum in the App Note. So, if I need to be precise, how can I make a good and cheap inductor for this purpose?

Why? can't I never make the circuit efficient enough so that the FET would run cool?

 

[4pyros]

You are missing some of the fundementals. You need to study some about buck converters.
Start Here;
Buck converter - Wikipedia, the free encyclopedia
Andy

 

[thecritic]

changing design

I couldn't understand, how the Original (microchip) Comparator Constant Current source worked.
The scenario is like this.
The -ve Vin constant.
The +Vin is below -Vin when the FET is fully off, and It is above -Vin when the FET is fully on.
So, at first the +Vin is Below -Vin, Output is Low and so the FET switches on.
As soon as Current builds-up enough, then +Vin will rise just slightly above -Vin, and FET should switch off. Then the current decrease, and +Vin will decrease slightly. Shouldn't this immediately switch on the FET again? Or is there some hysteresis in the Comparator and that determines the frequency?

I am also considering doing this instead, (since finding or making right size inductor seems too hard!) I will control the Duty Cycle of the PWM output to control the LAMPS brightness.

**broken link removed**

 

[4pyros]

I am also considering doing this instead, (since finding or making right size inductor seems too hard!) I will control the Duty Cycle of the PWM output to control the LAMPS brightness.

What exactly are you trying to do?

 

[ronv]

That should work as a lamp dimmer or flasher depending on the frequency of the pulses comming in.
I would change R1 to a 1k and r2 to a 470 ohm to speed up the FET turn on time.

 

[thecritic]

That should work as a lamp dimmer or flasher depending on the frequency of the pulses comming in.
I would change R1 to a 1k and r2 to a 470 ohm to speed up the FET turn on time.

Wouldn't it work by the same logic as a battery charger too? Thanks for the suggestion to change the Resistors.

 

[4pyros]

Wouldn't it work by the same logic as a battery charger too?

are you trying to make a PIC controlled charger? or a stand alone charger?

 

[crutschow]

Wouldn't it work by the same logic as a battery charger too? Thanks for the suggestion to change the Resistors.

No. The lamp has a resistance to limit the current. The battery has very low resistance and will not limit the current. That's why you need the inductor. It smooths out the current pulses and limits the peak current.