Obtaining dynamic voltage amplification from ARM chip

[blewisjr]

I know the chip can not do this itself.  I am trying to figure out the most accurate way to achieve my goal by regulating with a arm chip.  The reason this needs to be dynamic is due to the various variables which effect the output.  So here is an example.

Say the user sets the device to output 30 watts.  They are on a fresh battery which outputs 4.2 volts.  The ext e rnal resistance is 1.5 ohms.  So maintain the 30 watts wanted we need sqrt(30 * 1.5) = 6.7 V.  Which means we need to amplify by 2.5 V.  Now say the battery drains to 3.7 V output.  Now i need to amplify by 3V etc...

My first thought was PWM but not sure if the accuracy would be easy to maintain.  Any thoughts?

[dannyf]

Forget about "dynamic" for a second, what exactly are you trying to amplify?

[ElektroQuark]

It looks like he want to maintain constant output current.

[blewisjr]

Need to hold a constant output wattage based off of a fixed resistance.  The currenent provided should be constant as you are varying the voltage instead.

[dannyf]

So, you want a fixed output on a variable input?

Why not just generate a fixed output then? What's the point of providing an input to something when it does not change the output?

[blewisjr]

The wattage is settable from the device and the resistance is adjustable by building a different coil.  The throughput of current and voltage is comming from a Li-MN battery cell.  The chip needs to regulate the output based off the configuration and change in battery voltage.  As well as monitor safty reading to prevent thermal venting in case someone would be dumb enough to have settings that would exceed the constant discharge current of the battery.  It is technically a regulated device as if you just ran straight battery the wattage would drop as the battery drains.  The purpose of the device is to keep consistency.  So it is dynamic in many ways not fixed.

Here is a example of what I am shooting for.http://www.evolvapor.com/shop.php there are datasheets on the shop page more of a capability sheet then a data sheet.  Might give a more clear understanding.

[Zad]

I think we need to know a bit more about the power source and power sink that you intend driving. You can't usually set both current and voltage (and thus power) as the load derives one from the other. You would need to monitor the current draw, and perhaps use the PWM to set an output voltage which the output regulator uses as a control loop reference. Use software to monitor the total power and adjust the voltage to attain your set point. Then you just hope that your load is linear, because if it isn't you will probably end up with an oscillating output with the supply trying to chase a level that happened <x>ms ago.

[blewisjr]

To my knowledge the current is going to be a derived after effect since during use the resistance is stationary. P = ((V*V)/R).  The power source is a single 18650 IMR Sony VTC4 battery capable of a 30 A continuous discharge and a nominal voltage of 3.7 V.  At full charge it sits at 4.2 V and min discharge voltage of 2.5 V which you should never do as it can kill the battery for good.

[dannyf]

The wattage is settable from the device and the resistance is adjustable by building a different coil.

When facing a difficult problem, it is advisable to simplify it to a point where you can articulate it to a lay person (me).

What you want, it seems to me, is a controller that will output a given (user configurable?) amount of power in to a load.

Depending on the nature of the load, and your goals, it can be very difficult to do; it can be also very simple to do.

For example, if the load is fixed, all you need is to output a fixed (user configurable) voltage onto that load. It can be done entirely in analog (a linear solution); or it can be done digitally (like a dc/dc converter).

If the load is reactive, it can be quite difficult.

The key is to dumb it down.

[nctnico]

To my knowledge the current is going to be a derived after effect since during use the resistance is stationary. P = ((V*V)/R).  The power source is a single 18650 IMR Sony VTC4 battery capable of a 30 A continuous discharge and a nominal voltage of 3.7 V.  At full charge it sits at 4.2 V and min discharge voltage of 2.5 V which you should never do as it can kill the battery for good.

What you are looking for is a step-up or sepic converter. You can control the output voltage by having the controller look at the current and voltage and do some math.

[blewisjr]

The load resistance is fixed during operation to actually change the resistance you need to build a new wire coil of a different resistance to place in the atomizer.  The combined resistance of the coil plus atomizer is the load that is actually driving the current.  The goal here is to stabalize the power output over a period of battery drain time to keep consistency.  I believe the DNA 30 I linked earlier is using dc/dc I am not sure.  I was thinking along the lines of what nc said where I can monitor the voltage crunch and step up to keep the set wattage but I am not sure.  The issue I am having a hard time grasping is how to step up the voltage the required amount without frying the chip.

[nctnico]

The easiest way is to use a standard step-up IC. If the resistance is constant and known then you basically need a constant current. There are step-up converters intended for driving leds which are optimised for constant current. In the past I have used a chip from Linear Technology which had an input pin with which the current limit could be set by a variable voltage. I used the PWM output (and an RC filter to filter the PWM frequency) of a microcontroller to set the current and therefore the brightness of the LEDs. In your application it wouldn't hurt to measure the voltage as well and change the current if necessary in a software control loop.

[blewisjr]

Would chips like this be able to handle step ups form say 3.7 V to 8+V.  The issue here would be what can the circuit withstand because some of these setups may have pretty low resistance setups.  Like sub 1 ohm which really drives up that constant current.  For instance at 3.6V a 0.5 ohm setup is 7 amps.  My straight battery setup without the regulation can easily push over 20 amps.  Hence the use of a vtc4 battery.  In a 50 watt config this device would need 14 amps or so constant current at 3.6 volts.

[nctnico]

For those currents you have to look into a step-up converter with external mosfets and synchronous rectification. Ofcourse you'd need the proper inductors as well.

[blewisjr]

Thanks for all the info.  I have been further looking through things and it looks like I am going to need a Switching DC to DC converter of some sort to really even consider this kind of power drive.  Anyone know of any reliable IC's that do this so I can start browsing around and have a idea what I am looking for?

[tggzzz]

Anyone know of any reliable IC's that do this so I can start browsing around and have a idea what I am looking for?

Be aware that there are many subtle pitfalls with building your own switching power supplies.

Seriously consider buying a ready-made module, unless large-scale production cost is of paramount importance. (And if that is the case, you might like to consider the time and expense of your learning experience!)