12V Power Supply

[NavMicroSystems]

I have got the following challenge:

Provide stable 12V (max load 5 A)

from a varying input voltage of 10 to 15V.


I know I could use a switching regulator to build a SEPIC converter.
Or combine an integrated StepUp and StepDown converter
to do the Job.

BUT:

A PIC could certainly do the same (or a better) job.

Does anyone have a recommendation for the hardware/software design?

rgds

[Squibcakes]

Yeah, I recommend you scrap the PIC idea and use a switching regulator.. They were designed for this type of job.

Maxim have some great switching chips you can get for free.


;)

[G8RPI]

Hi,
This is not a trivial problem. The best way to do it is a step-up to higher voltage than your highest input, followed by a step down to 12V. To keep reasonable efficency use a high (I'd suggest 24-30V) intermediate voltage. This minimises currents and I2R losses.
A Pic is not the way to go. The other question is what is your load? does it really need 12V? If it is a piece of equipment does all of the circuit need 12V or will some run unregulated?
I've had a number of cases were people went to great lengths to provide the "correct" input voltage to a unit (and failed hence my involvement) when it was not needed. One example was a computer that was going to be fitted to an aircraft with a "28V" supply and the computer was specified at 24V. The 28 to 24V regulator dropped out during operation. However taking the cover off the computer revealed a packed DC-DC converter with an 18V - 32V input range!

Robert G8RPI.

[NavMicroSystems]

Robert,

the Load is a "Blackbox" that requires a stabilized 12V supply voltage. (max. 5A)

Power is provided by a solar panel and Lead batteries
(this is why the input voltage might be between 10 and 14 Volts)


There is no way to modify the "Blackbox" and tests have shown that it does not operate at a voltage lower than 11.5V.
The datasheet says the supply voltage must not be higher than 12.5 V.
I have not tested at what voltage higher than 12.5V it starts to generate smoke.

My initial thought was to use a StepUp converter in the first stage and then step down to 12V.

A second point is:
The environment temp will be up to 45 or even 50 deg C.
This requires the converter to work at high efficiency to reduce thermal losses.

Do you have any hints ?

rgds

[anj]

Gday
I have been working on exactly the same problem ( but only 1A ), for powering a small surveillance video camera from a battery ( attached to a telecope )
My input range is 14V down to 10 as the battery flattens and i need a constant 12V +/- 0.1V.
I have indeed gone down the DC-DC stepup, then linear stepdown ( LDO ) route, as space is at a premium, but if you have room, a DC-DC flyback system may be better.
Another advantage of the stepup then step down is the LDO stepdown unit will act as an output filter.
With my current setup, i can average around 76% efficiency across the bulk of the range. Your main problem will be when yr voltage exceeds say 12.3V, as the stepup unit stops switching, and all heat loss comes via the 12V stepdown. The higher yr input voltage, the better a heatsink you may require on that unit.
Andrew

[G8RPI]

Hi Again,
sorry for the slow response but my original reply was through the listserver and it seems a bit slow at the moment.
12V at 5A is 60W, quite a lot of power. DC-DC converters at this level are not very common, Power-one (formerly Melcher) do a 50W (12V 4A output) unit with 8-15V input but it costs £180.
This power would be pushing the limits for a simple Buck-Boost (stepup-stepdown) solution using the National or Linear Technologies devices. You are looking at a full-blown transformer based design. What is the average current draw of the unit and the duty cycle of the high power operation, 60W is a lot for solar power! What charge regulator are you using for your batteries? For long life you do not want to discharge them much below 11.5V anyway, so perhaps you colud switch in a simple dropper (3 silicon diodes in series) when you are charging and short it out with a relay when off charge. A Pic would make a good control for this and you could use more than one relay. A 13V (I don't think your unit will die the instant you exceed 12.5V) shunt clamp / regulator across your load could take care of any transient events without consuming power during normal operation.

Regards,
Robert.

[NavMicroSystems]

In the meantime I have solved the problem the following way:
(Thanks to Bob Blick)

[NavMicroSystems]

And here is (hopefully) the attachement:

[NavMicroSystems]

BTW

the effiency is more than 95% and a load of 8A is no problem.

[NavMicroSystems]

Attached you'll find the final version:

[NavMicroSystems]

And a picture of the prototype: