Analog out with High Power?

Hi Roger

If the output can be floating with respect to the input, a way of getting around the need for a SEPIC is to use an inverter. Yes the output voltage is negative with respect to the input. This way you do not encounter the buck/boost switchover problem and you should get slightly higher efficiency. SEPICs only give you about 70% efficiciency and have to have a low ESR cap between the 2 coils.

With an inverter, you use the negative output as the new ground. With battery chargers this can work quite well as long as the output does not have to be referenced to anything - as you now have 2 ground references.

You can inject a voltage into the feedback node of the dc/dc converter and fool the dc/dc into thinking it is out of regulation and tweak the voltage that way. However, you will only be able to adjust the output voltage down to the value of the reference.

Does any of this sound suitable?

Please let me know

Thanks

Bill Electronworks.co.uk - Electronic kits for education and fun

nd

e

hing

d

ould

ts

ot

in

Such

g for

Roger - What you are trying to invent is essentially a programmable 1.2-30V supply with current limiting/sensing in the 5A range. You might call it a battery charger but really it's a programmable power supply.

And then you layer on that it has to be powered from 13.8V. Charging batteries from batteries sounds a little weird, but maybe you mean it has to work from a car.

I think you've got a couple of choices if you want to build it from parts. Simplest is to go from 13.8V up to something in the 40V range with one converter, and then from 40V back to the charging voltage with a couple of selectable converters. This way, one half is always in boost, the other half is always in buck. The magnetics in the output converters will probably make the 1.2V converter not be the same as the 30V converter.

Alternatively, you could have one buck converter for 13.8V to loads under 9V, one SEPIC converter for 13.8V to loads from say 9 to 16V, and a third boost converter for 16V to 30V.

You will notice that any of the above solutions involves multiple sets of magnetics and circuitry.

Rather than re-invent the wheel, I think you would be most productive if you just got an off-the-shelf inverter followed by something like a B&K 1697 programmable supply. This will allow you to concentrate on your programming algorithms (sounds like you've got a lot of fancy-pants features you want) and less on the circuitry.

Tim.

Easy: Take a suitable SEPIC design and look at it's feedback input. It will typically reach regulation at some given FB voltage, for example

1.26V because that happens to be the bandgap voltage.

Buffer that path with an opamp and hook a, say, 1K resistor from there to FB. Now feed a current into that same FB node and gradually add current. Once you've added 1.26mA (or 1.26V via another 1K) your supply will have reached 0V. Set the other end to where you want the limit to be, i.e. 30V. You have to make sure that you'll never exceed the compliance range of that FB pin so check the data sheet. I've done this many times where I needed 0V to xxV from a fixed supply somewhere in that range.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

The inverter sounds interesting but there are also issues of voltage and current sensing A-D converters back to the microcontroller. Perhaps if I don't find a more straight forward approach, I could make a split supply and use op-amps to get the scale and offset in the A-D converters voltage range. I'm also wanting the ability to discharge batteries, etc..

Roger Neal

Yes, I need a variable switching power supply that I can adjust the voltage or current with a microprocessor or microcontroller.

The battery charger design I'm thinking of is intended for Radio Control models. With the rising cost of fuel, and the power they are getting from LiPo packs and brushless motors, electric models are getting more popular. Some of the battery packs are in the $300 range and many models are quite expensive. The ability to charge from a 12V battery or car in needed out at the flying field.

New LiPo packs are recommended to be broke in by not running them down as far for the first 5 or so cycles. I would like to be able to program the ability to break in new packs. Also, packs stored fully charged for a year lose about 20% of their capacity versus packs partialy charged lose 6% of their capacity. I would like the charger to be able to discharge to a storage voltage for the off season.

For now I would like to experiment for myself and see if I come up with anything good.

Roger - What you are trying to invent is essentially a programmable 1.2-30V supply with current limiting/sensing in the 5A range. You might call it a battery charger but really it's a programmable power supply.

And then you layer on that it has to be powered from 13.8V. Charging batteries from batteries sounds a little weird, but maybe you mean it has to work from a car.

I think you've got a couple of choices if you want to build it from parts. Simplest is to go from 13.8V up to something in the 40V range with one converter, and then from 40V back to the charging voltage with a couple of selectable converters. This way, one half is always in boost, the other half is always in buck. The magnetics in the output converters will probably make the 1.2V converter not be the same as the 30V converter.

Alternatively, you could have one buck converter for 13.8V to loads under 9V, one SEPIC converter for 13.8V to loads from say 9 to 16V, and a third boost converter for 16V to 30V.

You will notice that any of the above solutions involves multiple sets of magnetics and circuitry.

Rather than re-invent the wheel, I think you would be most productive if you just got an off-the-shelf inverter followed by something like a B&K 1697 programmable supply. This will allow you to concentrate on your programming algorithms (sounds like you've got a lot of fancy-pants features you want) and less on the circuitry.

Tim.

Thanks for the info thus far. The electronics part is readily available by many sources but I don't know how they are doing it. Here's some links to chargers similar to what I am wanting to work with. I just want to do my own software to add features not available and also add the ability to switch multiple packs in and out of the circuit.

Many of these chargers go up to 10A

Roger N

e by

to

my

A few points. Firstly car battery range is 10.8v - 15v, you'll need to cope with that lot for reliable operation.

2nd one of the advantages of SMPSUs is they act as transformers with automatically adjusting ratio. All you need do is make a step up 40v smpsu with current regulation, and it will autoadjust itself to whatever Vout is needed. You dont need external v control. Bob's your uncle.

NT

I like this idea very much, it's in line with what I had in mind. I'll try to set up a test circuit and see if I can get it to work. Thanks!

Roger N

try

to

. e 0V.

Hint: If you make this programmable make sure that your new supply always starts at 0V. Regardless of whether the software failed, froze up or the cable to the computer fell off.

Oh, and please bottom-post. It's easier to follow for others.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.