# Variable wattage module

#### boostedrst

Joined Feb 8, 2010
10
How can i make a module that will be able to adjust wattage by reading the voltage and resitance and compensating ? Say i have a one thing thats 2ohm 3.7v and another thats 5 ohm 3.7v and i want the module to maintain a set voltage .Also be able to ajust the wattage with a potentiometer .

I would like to do this with a maybe arm processor in a small package . Im not very good with these things . Can someone give me some help on doing this or post a circuit diagram that will do this ? Any help is greatly appreciated .

Normally i would use VV with a 2.5-6A boost or stepdown chip but variable wattage would be much better .

#### mcgyvr

Joined Oct 15, 2009
5,394
Wattage = voltage x current.
To say you want adjustable wattage is really incorrect (well it leaves it open to 2 choices).
You basically have 2 types of power supplies..Constant voltage and constant current.

In constant voltage you set the voltage and the amount of current is determined by the load but the voltage will stay stable. (power supplies don't just dump their full load current..a load like a resistor/lightbulb,etc.. draws or takes what it needs and leaves the rest)
In constant current you set the current level (like 20mA) and it will automatically vary its output voltage to ensure the current draw stays at what you set it at.

The common misconception with power supplies is for example a constant voltage 5V 1A power supply.. but they say "I only need 1/2A this won't work it will be too much"..That's just wrong. It just means that supply will only be supplying half of the total it "could" supply.

Now what do you really want.

#### boostedrst

Joined Feb 8, 2010
10
i do know that stuff variable voltage / current etc but i do mean variable wattage .

Heres a example of how it would be used .

Say you have vapor atomizers which is a heating coil that comes in various resistances .The lower the resistance the more power/heat/vapor/wattage obviously . The atomizers that come in different resistances are used on a 3.7v or Variable voltage battery device using a boost or stepdown chip if using stacked batteries .

The variable voltage device will take say a 3ohm atomizer and raise the voltage creating more vapor/watts thus pulling more current . Now you have to adjust the voltage to get the amount of watts you will like with each different resistance atomizer .With that you have to constantly adjust the voltage when changing a atomizer

A Variable wattage will read the resistance and adjust automatically to maintain a set wattage . So no matter what resistance atomizer is used it will maintain the same wattage .Now you can adjust the wattage so anything you run on it will maintain the user set wattage . This would be variable wattage and this is what im looking to build . It has been done before and it nothing relatively new . Im just looking for how to build it . I would like to use a arm processor .I am also looking to be able to adjust up to about 12W max . Power source is a 3.7v high drain battery

Last edited:

#### boostedrst

Joined Feb 8, 2010
10
Has anyone else done something like this before?

#### lightbeast

Joined Apr 6, 2012
7
Ok I know exactly what this guy is trying to do , Say you have a power supply and in the power supply you have a way to adjust voltage, but what you hook up to this supply comes in different loads and Ohms. Is there a way to regulate the wattage? Say I put on something with 3ohms then switch to something with 1.5 ohms with the new load I would have to re adjust my voltage and re calibrate it, but if it was regulated by watts and not volts it would have the same out put every time.

#### CDRIVE

Joined Jul 1, 2008
2,219
Quite surprising that this thread went dormant. For those that opine that too many topics are too basic this will be a challenge. If I was the OP I'd be looking at a uC approach.

#### #12

Joined Nov 30, 2010
18,217
There is no such thing as a wattage regulator chip. There are voltage regulator chips and current regulator circuits, but no wattage regulators. That makes this rather complicated, like microprocessor complicated. Fortunately, that is beyond my abilities.

Next helper?

#### crutschow

Joined Mar 14, 2008
26,977
Conceptually you would need to monitor both the output voltage and current of the supply, and adjust the output voltage until the watts (volts time current) is what you want. A microprocessor could do this, but you would need some form of loop compensation, such as PID, to make the control stable. If you weren't too concerned about speed you could just start at a low voltage and slowly increase it until the current and voltage are at the required level. A slow loop (lots of lag) makes it easier to stabilize.

#### pilko

Joined Dec 8, 2008
213
1 Enter power into b0 (manual power set point)
2 Measure V -- put into b1
3 measure I -- put into b2
wait 1
4 let b3 = b1/b2 (resistance of load)
5 let b4=b1*b1/b3 (power to load)
6 if b0>b4 then increase V
7 if b0<b4 then decrease V
wait 2

start:
8 Measure V -- put into b1
9 measure I -- put into b2
wait 1
10 let b4=b1*b1/b3 (power to load using new V and old R)
11 if b0>b4 then increase V
12 if b0<b4 then decrease V
wait 1

goto start

You may need some tuning

pilko

#### crutschow

Joined Mar 14, 2008
26,977
If you take pilko's program and and add a few "If-Then" statements that increases or decreases the voltage at a different rate depending upon how far the desired value is from the measured, you will have a simple Fuzzy Logic control loop. This improves the response time, similar to how a PID loop would.

#### CDRIVE

Joined Jul 1, 2008
2,219
1 Enter power into b0 (manual power set point)
2 Measure V -- put into b1
3 measure I -- put into b2
wait 1
4 let b3 = b1/b2 (resistance of load)
5 let b4=b1*b1/b3 (power to load)
6 if b0>b4 then increase V ' I think a window on these two
7 if b0<b4 then decrease V ' statements would be a in order
wait 2

start:
8 Measure V -- put into b1
9 measure I -- put into b2
wait 1
10 let b4=b1*b1/b3 (power to load using new V and old R)
11 if b0>b4 then increase V ' I think a window on these two
12 if b0<b4 then decrease V ' statements would be a in order
wait 1

goto start

You may need some tuning

pilko
Just a thought.

#### panic mode

Joined Oct 10, 2011
1,881
Rich (BB code):
5 let b4=b1*b1/b3 (power to load)
simpler and faster is P=V*I
Rich (BB code):
5 let b4=b1*b2

#### pilko

Joined Dec 8, 2008
213
@ crutschow and CDRIVE, ----Both good points.
Oscillations could be minimized with fast loop times, giving fast updates and small changes.

pilko

#### pilko

Joined Dec 8, 2008
213
@ panic mode, --- I used P=V^2/R to keep R constant and only use it once in the initial code. But, you're probably right and maybe that wasn't necessary.

pilko

#### #12

Joined Nov 30, 2010
18,217
Oddly, this makes sense to a person that learned how in 1975.
Is that actual code or just instructions to the OP?
I can do that! (with some effort)
What is this device you are programming for?
What is this language called?

#### pilko

Joined Dec 8, 2008
213
@ #12,

The language is BASIC. I use a version called PICAXE. The only programming hardware requirement is a cable.
It is so simple that I was writing simple code within a few hours.

pilko

#12

#### #12

Joined Nov 30, 2010
18,217
Oh! No wonder I can read it. Basic was one of my languages when the whole school shared an IBM mainframe to run our programs. I still have the books, but that's probably not bragging, as in, how much has Basic changed in 37 years?

This could be fun.

#### CDRIVE

Joined Jul 1, 2008
2,219
I recognized that code as Picaxe Basic immediately. The language is a Godsend for so, so many old seasoned techs & engineers that, at one time in their life, played with Quick Basic but never dove into uC's for fear of the dreaded 'C' code. The Programming Editor/Simulator is FREE, FREE, FREE and I love it!

#### CDRIVE

Joined Jul 1, 2008
2,219
@ #12,

The language is BASIC. I use a version called PICAXE. The only programming hardware requirement is a cable.
It is so simple that I was writing simple code within a few hours.

pilko
The Picaxe Editor/Simulator/Compiler is so slick that I had two simple programs written and simulated before the chips arived in the mail.

#### #12

Joined Nov 30, 2010
18,217
Big smile. It was so much fun to smoke the engineers that graduated in the 1950's by writing a Basic program in 1975. I can't believe that old language is still alive!

I can do this